Marginal and internal fit of 3D printed permanent crowns: effect of CAD/CAM software, preparation type and resin material Authors: Görkem Mümtaz, Beril Koyuncu Published: 2026 ABSTRACT AbstractBackgroundWith the advancement of digital technology, the stages of fabrication of temporary or permanent restorations with additive manufacturing (AM) are likely to become a routine in clinical dentistry. Compatibility with the teeth and surrounding gingival tissues is crucial for eliminating biological complications. Marginal and internal fit play a critical role in the mechanical continuity of the restoration and proper seating of the crown. The aim of the study is to evaluate the effects of CAD software, resin material and tooth preparation design on the marginal and internal fit of 3D printed resin nanohybrid crowns.MethodsThree typodont teeth were prepared with knife-edge, chamfer, and shoulder finishing lines. Crowns were designed using two CAD/CAM software systems (Exocad and 3Shape) and fabricated from two nanohybrid resin materials (VarseoSmile TriniQ and Saremco Crowntec). Marginal and internal fit were evaluated using the silicone replica technique (SRT) under ×40 magnification. Internal gaps were assessed separately for axial and occlusal regions. Data were analyzed using three-way ANOVA and Tukey post hoc tests (α = 0.05).ResultsPreparation design significantly affected marginal, axial internal and occlusal fit (p < 0.05). Material type influenced marginal and axial internal fit (p < 0.05), whereas CAD/CAM software affected only occlusal fit (p < 0.05). Marginal gap values in all groups were within clinically acceptable limits. Axial internal gaps were generally within reported acceptable ranges, while occlusal gaps were consistently higher and, in several groups, approached or exceeded commonly accepted thresholds.ConclusionPreparation design was the primary determinant of fit in 3D-printed resin crowns. While marginal and axial adaptation were clinically acceptable, occlusal internal discrepancies remain a critical limitation. Optimization of digital design parameters and preparation geometry is essential to improve the clinical performance of additively manufactured restorations.Keywords: Resin crown, Internal gap, Marginal gap, 3D printing, Additive manufacturing Background The clinical performance and long-term durability of restorations are strongly influenced by the accuracy of marginal and internal fit of the crowns [1, 2]. Inadequate marginal fit may promote plaque accumulation, microleakage, secondary caries, periodontal inflammation, and cement dissolution [2, 3]. Furthermore, inadequate internal adaptation of coping may result in increased cement thickness, which can adversely affect the mechanical stability of the prosthesis [4]. Therefore, achieving optimal marginal and internal fit remains one of the most critical objectives in contemporary prosthodontic treatment. In recent years, dentistry has experienced substantial advancements through the incorporation of digital technologies into clinical and laboratory workflows. Digital dentistry comprises a broad spectrum of technologies and applications, including three-dimensional (3D) imaging systems, computer-aided design and computer-aided manufacturing (CAD/CAM), 3D printing, intraoral scanning devices, AI-assisted (artificial intelligence) diagnosis, guided implant surgery, laser-assisted procedures etc. [5]. According to their manufacturing principles, CAD/CAM systems are categorized into subtractive manufacturing (SM) and additive manufacturing (AM) technologies. Although SM offers the advantage of an ongoing production process, its manufacturing accuracy may be compromised by wear of the milling burs during fabrication [6]. Material waste and limited capability of fabricating complex geometries are the main disadvantages of SM technology [7]. As AM (3D-printing) technology continues to evolve, reductions in cost and improvements in availability are expected to facilitate its broader utilization varying dental practices. Consequently, 3D printing may contribute to more individualized and efficient patient care by enabling the in-house fabrication of customized dental appliances with reduced production costs and shorter processing times [5]. Among the currently available AM systems, Digital Light Processing (DLP) technology has emerged as one of the most widely utilized methods for the fabrication of dental restorations layer-by-layer with a relatively short production time [6, 8]. Metals, ceramics and polymers are mainly used in 3D-printing [9].The resins employed in 3D-printing systems have similar composition with direct resin composites including methacrylate monomers, such as Urethane Dimethacrylate (UDMA), Bisphenol A-Glycidyl Methacrylate (Bis-GMA), Ethoxylated Bisphenol A Dimethacrylate (Bis-EMA), and Triethylene Glycol Dimethacrylate (TEGDMA) with photoinitiators of different wavelengths [10]. These materials have been reinforced with ceramic fillers, such as silica, aluminosilicates, or barium-containing particles and have traditionally been considered suitable for provisional restorations; however, more recently, manufacturers have introduced modified formulations designed for improving clinical service. Such advancements seek to broaden their indications from interim applications to definitive restorations [11]. Having less inorganic filler amount than block and disc composites, these printable materials must preserve adequate flowability and structural stability throughout the printing process to ensure accurate fabrication and reliable clinical performance [12]. At present, only a limited number of composite-based materials are available for the fabrication of 3D printed indirect definitive restorations. While certain products, such as Crowntec (Saremco), are described as composites, others, including VarseoSmile Crown Plus (Bego), are categorized as hybrid composites [13]. Post-processing procedures play a critical role in eliminating fitting inaccuracies in additively manufactured restorations. Several factors have been reported to influence the accuracy of objects fabricated using AM technologies, including laser speed, laser intensity, build orientation, layer number, the software employed, polymerization shrinkage, the quantity of support material, and post-processing protocols [14]. Various methods have been proposed for the assessment of marginal and internal adaptation of dental restorations, such as optical microscopy, profilometry, scanning electron microscopy (SEM), the silicone replica technique (SRT), the triple-scan method, microcomputed tomography (micro-CT), and 3D laser scanning [15]. The variability in measurement accuracy reported among studies may be attributed to the absence of standardized protocols, differences in laboratory conditions, and the diverse evaluation methods employed [16]. Among these methods, the silicone replica technique is frequently used due to its non-damaging nature [17]. This method is an accurate, repeatable and reproducible technique which can be used in both in vivo and in vitro conditions [18]. Moreover, previous studies have demonstrated that internal adaptation may vary according to anatomical region [19]. There are various studies which evaluated internal adaptation by focusing on different anatomical regions [1, 15, 17, 19–22]. Therefore, region-specific evaluation of internal fit may provide more clinically relevant information than a single overall internal gap value. Although the use of definitive 3D-printed nanohybrid resin crowns has recently increased, evidence regarding their marginal and internal adaptation remains limited and inconsistent. Furthermore, there is a lack of studies simultaneously evaluating the combined effects of preparation design, CAD/CAM software, and material type within a fully digital AM workflow. Therefore, the aim of the present study was to evaluate the marginal, axial internal, and occlusal internal fit of 3D-printed nanohybrid resin crowns fabricated using different preparation designs, CAD/CAM software systems, and resin materials. The novelty of this study lies in its region-specific analysis of internal adaptation (separately evaluation of axial and occlusal fit) and the simultaneous investigation of multiple clinical and digital variables within an AM workflow. The null hypotheses of the study are listed as follows: H0.1: Preparation design would have no effect on marginal, axial and occlusal fit. H0.1: Preparation design would have no effect on marginal, axial and occlusal fit. H0.2: CAD/CAM software would no effect on marginal, axial and occlusal fit. H0.2: CAD/CAM software would no effect on marginal, axial and occlusal fit. H0.3: Material type would have no effect on marginal, axial and occlusal fit. H0.3: Material type would have no effect on marginal, axial and occlusal fit. Methods Components and brand names of the materials used in this study are described in Table 1. Two different nanohybrid resin materials were used as subgroups: Crowntec (Saremco Dental AG, Rebstein, Switzerland) and VarseoSmile Triniq (Bego, Bremen, Germany). Additionally, these subgroups were classified according to three different finishing line designs (shoulder, knife-edge, chamfer) and two different CAD/CAM software systems (Exocad DentalCAD; Exocad GmbH and 3Shape Dental Designer; 3Shape A/S, Copenhagen, Denmark) as well. Table 1Components and brand names of the materials used in the study3D Printing Resin MaterialCompositionCompanyCrowntec [8, 23]Bis-EMA. Up to 50% in weight, 0.7-mm barium aluminum borosilicate glass fillers, 4,4′-isopropylidiphenol, ethoxylated and 2-methylprop-2enoic acid, silanized dental glass, pyrogenic silica, catalysts, inhibitors.Saremco Dental AG, Rebstein,SwitzerlandVarseoSmile Triniq [24]4,4′-isopropylidenediphenol, ethoxylated and 2-methylprop-2-enoic acid, Benzene acetic acid, alpha–oxo- methyl ester, diphenyl (2,4,6-trimethylbenzoyl) phosphine oxideBego, Bremen, Germany Components and brand names of the materials used in the study Saremco Dental AG, Rebstein, Switzerland The sample size for the study was determined through a power analysis conducted using G*Power Software (version 3.1.9.2) (Düsseldorf, Germany). For the evaluation of marginal and internal adaptation with an effect size (d) of 0.25, a power of 0.95 and an alpha level of 0.05, the minimum required sample size was calculated to be 10 per group. Resulting in 12 experimental groups, a total of 120 3D-printed models were planned to be fabricated with 40 models for each preparation type. To serve as the reference model for restoration production, three maxillary typodont right first molars (ANA 4; Frasaco, Tettnang, Germany) were prepared by the same researcher (G.M) with three different finishing lines: shoulder, chamfer and knife edge. Pin burs (Ökodent, 882-D016 C, 881D-016 C, Germany) were used in the preparations to ensure uniform step thickness and consistent reduction on both the occlusal and axial walls. A thickness of 1 mm was achieved in the shoulder-type preparation, while a 0.6 mm was obtained in the chamfer-type preparation. During tooth preparation, a range of 1.5–2 mm was preferred as the amount of occlusal reduction and 1–1.2 mm of reduction on the other walls were performed. The taper angle of the preparation was set between 6 and 10 degrees [25]. Although reduction values are presented as ranges based on clinical guidelines, all preparations were performed in a standardized manner by a single operator using depth-guiding burs to achieve uniform reduction across specimens. The prepared teeth were scanned using the 3Shape E4 desktop scanner (3Shape, Copenhagen, Denmark) and the obtained data were used to fabricate 3D printed tooth models (Model Resin, Form 4, Germany) with a 50-micron printing resolution using a 3D printer (Phrozen Sonic Mini 8 K, Taiwan). The master models were scanned using an intraoral scanner (TRIOS 4, version 1.7.6.1, 3Shape, Copenhagen, Denmark) following the manufacturer’s recommended protocol. The scanning was performed by the same operator under consistent lighting conditions, adhering to the scanning strategy suggested by the manufacturer. The data were then transferred to ExoCad (ExoCad GmbH, Darmstadt, Germany) and 3Shape (3Shape Dental Designer, 3Shape A/S, Copenhagen, Denmark) software programs for the design of the crowns. During the designing process in both software programs, the cement space was set to 60 microns, the additional cement gap to 10 microns, and the margin angle to 60 degrees. While designing, the crowns were adjusted to have a material thickness of 1 mm on the axial walls and approximately 1.7 mm within clinical limits on the occlusal surface. Figures 1 and 2 show the examples of the crown designs for chamfer preparation performed on 3Shape and ExoCad software programs respectively. Image of the crown designed for chamfer preparation with 3Shape software Image of the crown designed for chamfer preparation with ExoCad software For fabricating the 3D printed crowns, a common 3D printer recommended by both resin nanohybrid material manufacturers was preferred (ASIGA MAX UV, ASIGA, Australia). The cleaning and curing protocols were followed according to the respective recommendations of the two different manufacturers. For the Saremco Crowntec printing process, the base plate thickness was set to 0.3 microns, and the printing resolution was 50 microns. After printing, the models were cleaned using Saremco Aftergel (Swiss, Zahnärzten) liquid. For the curing process, the wavelength of the light was set between 320 and 500 nanometers, and curing was performed with 2 × 2000 flashes. For the VarseoSmile Triniq (Bego, Germany), the layer thickness was set to 50 μm. The printing orientation was standardized at 0 degrees to ensure optimal dimensional accuracy. Post-curing was carried out using a light-curing unit with a wavelength range of 320–500 nm, in accordance with the manufacturer’s instructions, using 2 × 1500 light flashes. Figure 3 exhibits the representative image of printed crowns on the build platform. Representative image of printed crowns on the build platform after manufacturing To assess the marginal and internal gap values, the SRT was used. For this technique, light-body (Elite HD; Zhermack, Rovigo, Italy) and heavy-body (Elite HD; Zhermack, Rovigo, Italy) additional type polyvinyl siloxane (PVS) impression materials were used. To ensure standardization in the process, the low-viscosity PVS material was homogenized using an automatic mixing device (Dispenser D2; Zhermack SpA, Italy) and injected into the inner surface of the crown restorations. To simulate the clinical cementation process, the crowns were seated on the prepared teeth with an average finger pressure. Following the manufacturer’s instructions, polymerization of the low-viscosity PVS was allowed to proceed for 2.5 min, after which the material was separated from the crown while remaining on the preparation surface. To support the low-viscosity PVS and replicate the interface accurately, high-viscosity PVS with a different color was applied on the external surface. Once the polymerization was complete, the silicone replicas were carefully removed as a single piece from the prepared surface (Fig. 4). If any tears or deformations were observed, the procedure was repeated to eliminate any potential errors. Separation of the master model from the impression material [Light-body additional type PVS (Elite HD; Zhermack, Rovigo, Italy)] Two adjacent incisions were made in the buccolingual direction along the sagittal plane of the silicone material and in the mesiodistal direction to obtain sectioned silicone replicas using a scalpel (No. 11, Plusmed, Turkey) (Fig. 5a and b). The surfaces of the obtained samples (Fig. 5c and d), including the cut faces, were visualized using a stereomicroscope (NIKON SMZ745, Nikon, Japan) (Fig. 6) and images were captured with an attached camera. For the marginal fit assessment of crown restorations, measurements were taken by identifying three points on each surface (mesial, distal, buccal, lingual). A total of 1,440 measurements were performed to evaluate the marginal fit. For the internal fit of a single crown restoration, a total of 3,600 measurements were taken across all 12 groups. These internal fit measurements were conducted on both the occlusal and axial (lateral) walls (mesial, distal, buccal, lingual). Internal fit measurements were evaluated separately for axial and occlusal regions due to their distinct anatomical configuration and clinical relevance. a-b Schematic representation of the buccolingual incisions and in the mesiodistal direction on the silicone replica. c-d Representative images of the incisions An example of gap measurement using stereomicroscope In the evaluation of the findings obtained in the study, IBM SPSS Statistics 22 (IBM Corp., Armonk, NY, USA) software was used for statistical analyses. While analyzing the study data, the normality of the parameters was assessed using the Kolmogorov-Smirnov and Shapiro-Wilk tests, and it was determined that the parameters showed a normal distribution. To evaluate the combined effect of preparation design, material, and CAD/CAM software system on marginal fit as well as the internal gaps of the axial and occlusal walls, a three-way ANOVA test was applied, followed by Tukey HSD test as the post hoc analysis. Statistical significance was considered at the level of p < 0.05. Results Evaluation of the effect of preparation type, material, and software system on marginal fit Preparation design had a significant effect on marginal fit (p < 0.05), with knife-edge preparations generally showing lower marginal gap values compared to chamfer and shoulder designs. Material type also significantly influenced marginal adaptation (p < 0.05), whereas CAD/CAM software had no significant effect (p > 0.05). No significant three-way interaction between preparation design, material, and software was observed for marginal fit (p > 0.05). Mean ± standard deviation of marginal gap values are shown in Table 2. Table 2Mean ± standard deviation of marginal gap valuesSoftware SystemChamferShoulderKnife EdgeMaterialsMean ± SSMean ± SSMean ± SS 1 p 3ShapeTriniq99,58 ± 10,09a110,05 ± 9,11a82,29 ± 11,87b0,001*Crowntec96,92 ± 9,86a120,79 ± 14,43b85,82 ± 8,01a0,001*2p0,5590,0620,446ExoCadTriniq103,12 ± 11,44a100,07 ± 21,49a80,97 ± 9,43b0,006*Crowntec103,21 ± 12,74a116,56 ± 8,96a104,74 ± 44,66a0,4982p0,9870,045*0,1173Shape-ExoCadTriniq 3p0,4720,2010,786Crowntec 3p0,2330,4410,204Three-way ANOVA Test; *p < 0,051p: p values between preparation types2p: p values between materials3p: p values between software systemsDifferent superscript letters in the columns indicate differences between preparation types Mean ± standard deviation of marginal gap values Three-way ANOVA Test; *p < 0,05 1p: p values between preparation types 2p: p values between materials 3p: p values between software systems Different superscript letters in the columns indicate differences between preparation types Evaluation of the effect of preparation type, material, and software system on the axial internal gap Preparation design and material type significantly affected axial internal gap values (p < 0.05). In contrast, CAD/CAM software alone did not show a significant effect (p > 0.05). Although some interaction effects were statistically significant, these were limited and interpreted cautiously. Overall, axial internal gaps were relatively consistent across groups. Table 3 shows the mean and standard deviation of internal axial gaps. Table 3Mean ± standard deviation of internal gap values on the axial surfacesSoftware SystemChamferShoulderKnife EdgeMaterialsMean ± SSMean ± SSMean ± SS 1 p 3ShapeBego Triniq117,95 ± 24,4a126,34 ± 15,25a107,20 ± 7,31a0,060Crowntec114,45 ± 11,19a151,59 ± 21,4b101,02 ± 10,66a0,001*2p0,6870,007*0,148ExoCadBego Triniq111,1 ± 16,11a110,22 ± 20,98a120,68 ± 20,9a0,423Crowntec145,49 ± 22.76a106,33 ± 11.47b125,89 ± 19,93ab0,001*2p0,001*0,6130,5753Shape-ExoCadBego Triniq 3p0,4690,0650,080Crowntec 3p0,002*0,001*0,003*Three-way ANOVA Test; *p < 0.051p: p values between preparation types2p: p values between materials3p: p values between software systemsDifferent superscript letters in the columns indicate differences between preparation types Mean ± standard deviation of internal gap values on the axial surfaces Three-way ANOVA Test; *p < 0.05 1p: p values between preparation types 2p: p values between materials 3p: p values between software systems Different superscript letters in the columns indicate differences between preparation types Evaluation of the effect of preparation type, material, and software system on the occlusal internal gap Occlusal internal gap values were significantly influenced by preparation design and CAD/CAM software (p < 0.05), while material type had no significant effect (p > 0.05). Compared to axial regions, occlusal gaps were consistently higher across all groups. Although statistically significant interaction effects were observed, their interpretation was limited to avoid overemphasis on multiple comparisons. Table 4 exhibits the mean and standard deviation of occlusal gaps. Table 4Mean ± standard deviation of internal occlusal gap valuesSoftware SystemChamferShoulderKnife EdgeMaterialsMean ± SSMean ± SSMean ± SS 1 p 3ShapeBego Triniq151,86 ± 48,37a157,5 ± 31,13a145,38 ± 11,97a0,729Saremco130,46 ± 19,37a191,86 ± 52.6b151,26 ± 24,16a0,002*2p0,2110,0920,502ExoCadBego Triniq147,76 ± 21,46a188,83 ± 51,38b158,14 ± 12,53ab0,026*Saremco182,12 ± 48,15a197,07 ± 47,9a161,89 ± 29,85a0,2022p0,0540,7150,7203Shape-ExoCadBego Triniq 3p0,8090,1160,032*Saremco 3p0,006*0,8190,393Three-way ANOVA Test; *p < 0,051p: Significance values between preparation types2p: Significance values between materials3p: Significance values between software systemsDifferent superscript letters in the columns indicate differences between preparation types Mean ± standard deviation of internal occlusal gap values Three-way ANOVA Test; *p < 0,05 1p: Significance values between preparation types 2p: Significance values between materials 3p: Significance values between software systems Different superscript letters in the columns indicate differences between preparation types Discussion According to the findings of the present study, H0.1 was rejected and H0.2 was rejected for occlusal internal fit but failed to be rejected for marginal and axial fit. H0.3 was rejected for marginal and axial fit but failed to be rejected for occlusal fit. Although statistically significant interaction effects were observed, their clinical relevance should be interpreted with caution. Greater emphasis was placed on main effects and clinically meaningful differences. In the present study, instead of calculating an overall internal gap value, axial and occlusal internal gaps were evaluated separately due to their distinct biomechanical roles and morphological differences. In some groups there were relatively higher standard deviation values especially in occlusal regions. This may be attributed to the complex morphology of the occlusal surfaces and the measurement technique used. Similar to our study, there are many studies which preferred to measure the internal axial and occlusal gap values separately [1, 17, 19–22, 26]. There were diversities in marginal gap measurements particularly in knife-edge configurations as well. The sensitivity of the SRT in thin marginal areas and less defined and uncertain finish line of the knifeedge preparation are thought to be the reasons for this variability. SRT was preferred in this study taking its advantages into consideration. In their study, Topdağı et al. found consistent gap values comparing silicon replica technique and Micro-CT analysis [2]. In addition to its advantages such as being non-invasive, low cost, simplicity, ease of application, and suitability for intraoral use, the SRT has several limitations as well. Difficulties in identifying the finish line accurately and the thin and elastic nature of the light-body silicone layer may compromise the precision of fit assessment. Furthermore, potential inaccuracies may arise during sectioning and measurement procedures, and the silicone layer may be susceptible to tearing during removal of the veneer from the prepared tooth surface [27]. For relevant evaluation, the number of measurement points must represent all anatomical regions and provide statistical validity. In 2004, Gassino et al. stated that a minimum of 18 measurements was required to achieve reliable results for single experimental crown gap [28]. In this study, 28 measurement points were used per crown. Different cement gap settings are currently used in AM and adaptation studies. Shin et al. evaluated various cement gap values in their study and recommended a 70-µm cement gap setting for optimal marginal and internal fit of 3D printed resin crowns [21]. Hence, a 70-micron cement space was used in both software programs in the present study. Studies have shown that printing at 0 degrees results in better dimensional accuracy and marginal fit [29, 30]. The build orientation has an influence on the marginal discrepancy of single crowns manufactured when using DLP [30]. Therefore, crowns in this study were printed at 0 degrees with DLP technology. The present study evaluated the marginal and internal fit of 3D-printed nanohybrid resin crowns fabricated using different preparation designs, CAD/CAM software systems, and materials. The findings demonstrated that preparation design was the most influential factor affecting both marginal and internal adaptation. Inadequate internal fit may adversely influence the seating of the restoration and marginal adaptation in a roundabout way [31]. Such misfits can promote microleakage and recurrent caries, increase dentinal hypersensitivity, contribute to periodontal complications, and ultimately compromise the longevity and clinical success of the restoration [31, 32]. Marginal adaptation is closely related to the accurate reproduction of the finish line, as excessive marginal discrepancies may result in cement dissolution and the development of secondary caries. Similarly, internal fit is influenced by the quality of tooth preparation, with smooth and well-contoured internal surfaces promoting proper seating and adaptation of the restoration [32]. In the present study, marginal gap values in all groups remained below the clinically acceptable threshold of 120 μm, which is consistent with previously reported and widely accepted clinical threshold marginal gap value in the literature [33]. This suggests that AM systems can produce clinically acceptable marginal adaptation under controlled conditions. However, different threshold values for marginal and internal gap have been reported [27]. Komine et al. considered marginal discrepancies ranging from 86 μm to 154 μm to be clinically acceptable [34], whereas Miura et al. reported that internal gaps between 41 μm and 141 μm were satisfactory [35]. It was stated that internal gaps exceeding 122 μm may have an adverse effect on the fracture resistance of restorations [15]. Much higher values like 200 to 300 μm were mentioned in a study by Aktaş et al. for internal gaps [36]. Nevertheless, no clear consensus has yet been established in the literature regarding the clinically acceptable values for marginal and internal adaptation of restorative infrastructures. Internal adaptation demonstrated region-specific diversity. The occlusal regions consistently exhibited higher internal gap values compared to axial walls. This finding is in agreement with previous studies reporting that occlusal gap being larger than the axial gap [21, 22, 37]. Higher occlusal gap may be attributed to the complex and irregular morphology of the occlusal surface, typically resulting in a thicker cement space compared with the axial region [38]. The increased occlusal gap values observed in this study may be attributed to the limitations of AM processes. Occlusal surfaces are more susceptible to inaccuracies in 3D printing, since the precise reproduction of cusp morphology may be compromised by the presence of support structures positioned on the occlusal aspect [39]. Material type significantly affected marginal and axial internal fit but did not significantly influence occlusal adaptation. AM is based on photopolymerization processes that depend on factors such as resin viscosity, layer thickness, building orientation, the uniformity of light exposure and post-processing stages. These parameters may contribute to dimensional inaccuracies [40]. Compared with Crowntec, VarseoSmile demonstrates greater dimensional stability owing to its lower organic matrix content and reinforcement with hybrid ceramic particles. In contrast, the higher organic phase content of Crowntec may increase its susceptibility to water sorption, potentially resulting in dimensional alterations over time [41]. Preparation design significantly influenced both marginal and internal fit. Similar to our study, Alharbi et al. reported favorable marginal and internal adaptation with knifeedge preparations with lowest internal discrepancy and vertical gap values in their study [42]. In contrast, Zarone et al. stated that there was no significant difference between vertical (undetectable finish line) and horizontal (shoulder) preparation designs [43]. Rounded shoulder with bevel exhibited the least absolute marginal discrepancy values in two different studies [42, 44] The lower marginal gap values observed in knife-edge preparations in the present study may be associated with the simplified marginal geometry and the way AM reproduces finishing line details. However, this finding should be interpreted with caution, as knifeedge preparation design may present clinical limitations in terms of margin detectability and periodontal response. Further studies are needed to clarify this relationship. The effect of CAD/CAM software systems was limited, with a statistically significant influence observed only in occlusal internal gap values. This suggests that software-related differences in occlusal morphology design may have a greater impact on complex surfaces than on simpler areas which are considered to be axial or marginal regions. Huong et al. evaluated Dentbird and Cerec InLab softwares with semi-automatic or automatic finish line detection methods and found out that there was no significant difference between the two softwares regarding finish line detection. Only automatic detection of Dentbird software was statistically significant from all the other groups and exhibited higher internal and marginal gap values [45]. In contrast, Rençber-Kızılkaya and Kara stated that the software programs of different CAD systems (Dentbird, Exocad, InLab) affect the marginal and internal fit of crowns produced [46]. Another study by Kim et al. evaluated the marginal and internal fit of crowns produced by three softwares (conventional: 3Shape; AI-based: Dentbird and Single-Pro). There were larger occlusal gap values in the AI-based software (Single-Pro) similar to the present study [47]. Taking into account all of these, the clinical impact of software selection appears to be less important compared to preparation design. A limitation of this study is that the SRT only provides two-dimensional, sectional analysis. Also, despite using standardized settings, exact anatomical matching across different software is not always feasible. Being an in vitro study, results reflect ideal conditions without patient variability. The use of a single operator improves standardization but may prevent drawing general conclusions as well. Conclusion Within the limitations of this in vitro study, preparation design was the most influential factor affecting both marginal and internal fit (axial and occlusal region). Knife-edge preparations demonstrated lower marginal gap values compared to other finish lines, although their clinical applicability should be interpreted with caution due to potential limitations in margin definition. Although marginal and most axial internal gaps were within clinically acceptable limits, occlusal discrepancies exceeded recommended thresholds in all groups, indicating a potential risk for increased cement thickness and compromised mechanical performance. From a material and software perspective, both evaluated systems demonstrated comparable performance in most parameters. From a clinical perspective, particular attention should be paid to occlusal internal adaptation when using 3D-printed resin crowns. When occlusal accuracy is of primary concern, the choice of CAD/CAM software may play a more critical role than material selection. Optimization of digital design parameters and preparation geometry is essential to improve the clinical performance of additively manufactured restorations. Key takeaways ✔ Preparation design is the primary determinant of fit. ✔ Preparation design is the primary determinant of fit. ✔ Occlusal internal adaptation is the most critical limitation in AM crowns. ✔ Occlusal internal adaptation is the most critical limitation in AM crowns. ✔ Marginal fit is generally acceptable. ✔ Marginal fit is generally acceptable. ✔ Software choice has limited clinical effect. ✔ Software choice has limited clinical effect. Acknowledgements Not applicable. Abbreviations Three-dimensional printing Computer aided design/computer aided manufacturing Subtractive manufacturing Additive manufacturing Digital Light Processing Scanning electron microscopy Silicone replica technique Microcomputed tomography Authors’ contributions G.M: Manuscript writing, data gathering, Study and experimental design, specimens preparation, performing the experiment B.K: Study and experimental design, data gathering, analysis and interpretation of the results, manuscript proofreading, manuscript editing. Both authors read and approved the final manuscript. Funding Not applicable. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.