Patterns and Mood Effects of Concomitant Polydrug Use of Ecstasy and Alcohol in Dutch and British Electronic Dance Event Visitors: An EMA Study Authors: Desirée B. Spronk, Ruben J.J. van Beek, Robin L.A. Smits, Rosa Andree, Wouter den Hollander, Tom P. Freeman, H. Valerie Curran, Jon Waldron, Meryem Grabski, Matthijs Blankers, Margriet W. van Laar Published: 2026 ABSTRACT AbstractIntroductionGiven the widespread prevalence of concomitant polydrug use of ecstasy and alcohol, it is important to better understand the nature and effects of concomitant use. The current study aimed to investigate patterns of concurrent ecstasy and alcohol use and their possible association with mood alterations.MethodsNightlife attendees who reported having used ecstasy three times or more in the past year were recruited and participated in an ecological momentary assessment protocol as part of a larger study on substance use in the European nightlife context (ALAMA study). Substance use and mood states were assessed on four timepoints per day or night with 2-h intervals on Friday and Saturday over a 5-week period.ResultsAnalyses of the course of alcohol consumption during an “ecstasy and alcohol” night (i.e., with at least one time ecstasy use) showed that alcohol consumption decreased in the hours following the first night-time use of ecstasy. Alcohol consumption increased more over time on “no ecstasy” nights than on “ecstasy and alcohol” nights. There were stronger improvements in positive mood states and a stronger decline in negative mood states over the course of “ecstasy and alcohol” nights compared to “no ecstasy” nights.ConclusionConcomitant ecstasy and alcohol use among nightlife attendees is common and is associated with specific use patterns and mood effects. These findings underscore the need for nightlife-specific harm-reduction approaches that consider the social context, patterns, and the reinforcing mood effects associated with combined ecstasy and alcohol use. Introduction The use of ecstasy (3,4-methylenedioxymethamphetamine or XTC) is common at festivals and electronic dance events in many countries worldwide. Among nightlife attendees, ecstasy is the most commonly used drug after alcohol and cannabis [1]. Ecstasy is also the most frequently used drug at festivals and outdoor parties [2]. Notably, the majority of people who use ecstasy combine ecstasy with alcohol [3–5]. Combined ecstasy-alcohol use poses greater risks compared to the risks associated with the use of either drug alone. Alcohol enhances the likelihood of ecstasy-related health issues such as hyperthermia, dehydration, and hyponatremia [6, 7]. This combination was also the most frequently reported at first aid stations during Dutch electronic dance events [8] and is linked to higher rates of aggression and sexual risk behaviour [5, 9]. Understanding co-use patterns is therefore essential. Previous studies on the co-use patterns of alcohol and ecstasy have indicated that people generally start with alcohol and progress to the use of other drugs, including ecstasy [10–12]. A more interspersed pattern of alcohol and ecstasy use was also reported, i.e., alcohol was used before, during, and after the use of ecstasy [13]. Most findings are based on retrospective questionnaires, which are prone to recall bias. One exception is an ecological momentary assessment (EMA) study in the USA, in which drug use patterns were assessed in real-time while people were in the nightlife setting. It was shown that alcohol consumption was more likely before compared to during and after ecstasy use [12]. However, the sample was small, and similar patterns in European nightlife remain unexplored. More broadly, little is known about the real-world timing of ecstasy-alcohol co-use, i.e., how alcohol consumption develops before and after ecstasy intake, and how these patterns compare to nights without ecstasy use. Comparing alcohol use trajectories on ecstasy-alcohol versus alcohol-only nights within the same individuals also offers a meaningful way to isolate ecstasy-specific changes in drinking behaviour from typical drinking patterns. Co-use of ecstasy and alcohol results in different acute mood effects compared with the effects of either drug alone. Ecstasy was found to counteract feelings of sedation and lethargy produced by alcohol [14–17]. Alcohol has also been found to prolong ecstasy-induced euphoria, suggesting it can enhance the positive mood effects of ecstasy [16]. These studies indicate that combined use is associated with decreased negative mood states and enhanced positive mood states. Whilst interactive effects on mood seem plausible, most studies were performed in controlled laboratory settings and do not take into account the dynamics of ecstasy and alcohol co-use behaviour in real-world settings. It is also unclear how mood effects develop over time in the course of a night out. Using EMA allows us to examine how these mood effects unfold in real-time and in naturalistic nightlife settings, providing insight into moment-to-moment mood dynamics that extend beyond what is expected based on pharmacological effects alone. We examined ecstasy-alcohol use patterns and associated mood effects using EMA. We had the following three research aims:1.To assess the timing of alcohol consumption on nights when ecstasy was used.2.To assess the quantity and course of alcohol use on “ecstasy and alcohol” nights compared to “no ecstasy” nights.3.To assess differences in mood effects between “ecstasy and alcohol” and “no ecstasy” nights. 1.To assess the timing of alcohol consumption on nights when ecstasy was used. To assess the timing of alcohol consumption on nights when ecstasy was used. 2.To assess the quantity and course of alcohol use on “ecstasy and alcohol” nights compared to “no ecstasy” nights. To assess the quantity and course of alcohol use on “ecstasy and alcohol” nights compared to “no ecstasy” nights. 3.To assess differences in mood effects between “ecstasy and alcohol” and “no ecstasy” nights. To assess differences in mood effects between “ecstasy and alcohol” and “no ecstasy” nights. Methods Participant Recruitment and Selection Eligibility criteria included being 18–36 years old, residing in the Netherlands or the UK, owning a compatible smartphone, and having used ecstasy at least three times in the past year. Participants were recruited from the Electronic Music Scene Survey (EMSS), part of the ALAMA nightlife project, and were invited by email if they had reported ecstasy use at least three times in the past year. The EMSS applied an inclusion criterion of attending a minimum of 6 electronic dance parties in the past year. Therefore, this criterion also applies to ALAMA participants. In total, 5,535 EMSS participants were invited for the EMA screening survey, of whom 1,176 did so. After excluding participants who did not meet the inclusion criteria, 410 participants were invited to take part in the EMA study. Of these, 273 participants completed the 5-week period (Fig. 1). Flowchart of the study population. Only participants who had used ecstasy at least once during the EMA study period were included in the analyses. From this group, we further selected participants who reported at least one night of alcohol use during the study period, resulting in a final sample of 136 participants. All participants provided written informed consent and received financial compensation of up to EUR 150 (EUR 25 per week), with an additional optional bonus of EUR 25 for completing more than 90% of the daily surveys. Procedure This study used EMA to measure drug use, mood states, and environmental circumstances. EMA enables repeated assessments of real-life drug use and momentary mood states over time [18]. Data were collected via a custom-made smartphone app between July and October 2018; methodological details are provided in the online supplementary Materials (for all online suppl. material, see https://doi.org/10.1159/000552706). Data Preparation In preparation for the analyses of “ecstasy and alcohol” nights, the four EMA timepoints were recoded relative to the moment ecstasy was taken. The timepoint at which ecstasy was taken was set to 0. Absolute timepoints were then scaled relative to the first and last reported moments of ecstasy use. When participants used ecstasy multiple times during an evening, all those timepoints were set to 0, including intermediate timepoints where no ecstasy use was reported (see Fig. 2 for an example). This approach was adopted because the exact timing of ecstasy consumption relative to EMA prompts could not be determined. Example of rescaling of timepoints. Absolute timepoints were scaled relative to the first and last moments of ecstasy use. The exact timing of ecstasy consumption relative to EMA prompts could not be determined. Thus, when participants used ecstasy multiple times during an evening, all those timepoints were set to 0, including timepoints in between where no ecstasy was taken. Participants’ locations during the night were dichotomised into nightlife versus home settings. Bars, pubs, cafés, parks, outside, clubs, and festivals were classified as nightlife settings, whereas all other locations (e.g., [a friend’s] home, house party, or work) were classified as home settings. Two datasets were used for the analyses. The first was the total dataset that contained both nights when ecstasy was used (“ecstasy and alcohol” nights) or when no ecstasy was used (“no ecstasy” nights), with a total of n = 781 nights (705 nighttime, 76 daytime). Across study days, participants completed, on average, 2.4 out of 4 EMA prompts per day. In the “ecstasy and alcohol” nights group, participants used ecstasy and alcohol, and may have used other drugs, but not necessarily. In the “no ecstasy” nights group, participants used alcohol and did not use ecstasy, but may have used other drugs. The second dataset was a subset of the total dataset that contained only “ecstasy and alcohol” nights, with a total of n = 168 nights (127 nighttime, 41 daytime) and an average of 3.1 completed EMA prompts per day. For a sensitivity analysis, the total dataset was used to differentiate two groups within the “no ecstasy” nights group: alcohol and other drugs used, and only alcohol used. Henceforth, these three groups will be described as an “ecstasy and alcohol” night, an “other drugs and alcohol” night, and an “alcohol only” night. In the “other drugs and alcohol” night group, drugs other than ecstasy were always used during the night. In addition, to isolate the effects of ecstasy use, nights during which other drugs were used were excluded, and the analyses were repeated comparing “ecstasy and alcohol” nights and “no ecstasy” nights. Data Analysis Descriptive statistics were used to determine the frequencies (%) or mean scores and standard deviations (SDs) of demographic variables, nightlife attendance, and drug use history. The primary aim of the present analyses was to examine overall within-evening trajectories in mood and how these differed between night types and relative to ecstasy intake. First, we assessed alcohol consumption relative to ecstasy use in the “ecstasy and alcohol nights” dataset by using a robust linear mixed model (robustlmm package in R) with alcohol consumption as the dependent variable and relative timepoint as the independent variable. Random intercepts were included to account for clustering of observations within participants and within days nested within participants (specified as [1 | participant/day]). Additionally, we repeated these models with the relative timepoint categorised into three groups (pre, at, and post ecstasy intake) and the absolute timepoint as the independent variable. Then, to assess the course of alcohol use on “ecstasy and alcohol” nights compared to “no ecstasy” nights, we first used a linear mixed model with alcohol consumption as the dependent variable, type of night (“ecstasy and alcohol” or “no ecstasy”) as the independent variable, and a random intercept (1 | participant/day). Next, we added an interaction term between the type of night and absolute timepoint to assess whether the course of alcohol consumption was different between the two groups. The same modelling steps were performed for all mood outcomes. All models were adjusted for country and location to check whether this affected the interpretation of the findings. Two sets of sensitivity analyses were conducted. First, we repeated the analyses of alcohol consumption and mood effects with three night types instead of two. Thus, we compared “other drugs and alcohol” nights and “alcohol only” nights with “ecstasy and alcohol” nights. Second, after exclusion of nights in which other drugs besides ecstasy were used, we repeated the comparison of alcohol consumption and mood in “ecstasy and alcohol” nights versus “no ecstasy” nights, as well as the interaction with type of night and absolute timepoint. For all analyses, p values were calculated using the degrees of freedom obtained from the equivalent canonical linear mixed model (lme4 package). No imputation was performed; mixed-effects models were estimated using all available observations. Results Sample Characteristics The total study population consisted of 136 participants who used ecstasy at least one time and drank alcohol during the study period. Regarding the concomitant alcohol and ecstasy use, 128 of the 136 individuals (94.1%) reported the combined use within the same night at least once during the study period (“ecstasy and alcohol” nights). The mean sample age of the total study population was 25.3 years (SD: 4.4; Table 1). The majority of participants (55.1%) identified as male. Of the total sample, 44.9% were Dutch, and 55.1% were British. Participants reported attending an average of 18.6 dance events in the last year (SD: 15.4). All participants reported having used both ecstasy and alcohol in the past year (online suppl. Table 1). Most participants reported using ecstasy every 2 or 3 months (42.1%) or monthly (36.5%). In addition, many reported the use of other substances at least monthly: cannabis (38.0%), cocaine (29.1%), amphetamine (11.7%), and 4-FA (14.9%). The subsample consisting of only “ecstasy and alcohol” nights did not differ substantially from the total population. Table 1.Baseline characteristics of the study populations​“Ecstasy and alcohol” and “no ecstasy” nights“Ecstasy and alcohol” nightsNights N = 781 N = 168Individuals N = 136 N = 128Age, years, mean (SD)25.3 (4.4)25.3 (4.3)Gender, n (%) Man75 (55.2)73 (57.0) Woman60 (44.1)54 (42.2) Other1 (0.7)1 (0.8)Country, n (%) The Netherlands61 (44.9)59 (46.1) UK75 (55.1)69 (53.9)Attended dance events in the last year, mean (SD)18.6 (15.4)18.7 (15.7) Baseline characteristics of the study populations Use of Other Substances during the Night The use of other substances in addition to alcohol and ecstasy was very common during “ecstasy and alcohol” nights (64.3% of all nights). In contrast, the use of other substances was less frequent on “no ecstasy” nights (28.1% of all nights). The substances most frequently used during “ecstasy and alcohol” nights were cannabis (14.9%), cocaine (19.1%), amphetamine (23.2%), and ketamine (19.1%) (online suppl. Table 2). A similar pattern was observed for “no ecstasy” nights, but with lower prevalence rates. Alcohol and Ecstasy Use on “Ecstasy and Alcohol” Nights Focusing on only the “ecstasy and alcohol” nights, 96 participants reported one night of concomitant ecstasy-alcohol use, 24 participants reported two such nights, and 8 participants reported three nights. Ecstasy use was most frequently reported at timepoint 4, followed by timepoint 3 (online suppl. Table 3). During most nights, ecstasy was taken once (44.6%). In 28.0% of the nights, ecstasy was taken twice, in 21.4% three times, and in 6.0% four times. Reliable data on ecstasy dosage were not available and were therefore not reported. The average number of alcoholic drinks during an “ecstasy and alcohol” night was 9.9 (SD = 9.8; range 1–64). Within these nights, most people consumed alcohol across multiple timepoints (see Fig. 3). Estimated marginal mean scores and 95% confidence intervals of alcohol consumption (drinks) at the different absolute timepoints on “ecstasy and alcohol” nights. Figure 4 shows the course of concomitant alcohol use relative to the moment of ecstasy use within a night. Alcohol consumption occurred mostly before and at the time of ecstasy intake. Analyses of alcohol use patterns relative to ecstasy intake confirmed a decrease in alcohol consumption following ecstasy use (Table 2). This decrease was most pronounced between the timepoint of ecstasy intake and the first subsequent timepoint (p < 0.001). Estimated marginal mean scores and 95% confidence intervals of alcohol consumption (drinks) relative to the moment of ecstasy intake. Timepoint 0 (at) is at or during ecstasy intake, and −1, −2, −3 (pre) are the timepoints before the intake, and 1 and 2 (post) are after the intake. Table 2.Course of alcohol consumption on “ecstasy and alcohol” nights​Estimate (95% CI) p valueRelative timepoints Intercept2.71– Timepoint −3−0.34 (−1.39, 0.72)0.53 Timepoint −20.08 (−0.43, 0.58)0.76 Timepoint −10.17 (−0.20, 0.53)0.37 Timepoint 1−1.42 (−2.10, −0.75)<0.001 Timepoint 2−1.16 (−2.92, 0.61)0.20Relative timepoints – pre, at, post Intercept2.71– Pre ecstasy0.12 (−0.20, 0.45)0.46 Post ecstasy−1.40 (−2.05, −0.75)<0.001Absolute timepoints Intercept2.68– Timepoint 20.34 (−0.04, 0.73)0.08 Timepoint 3−0.01 (−0.40, 0.38)0.96 Timepoint 4−0.56 (−0.95, −0.16)0.006 Course of alcohol consumption on “ecstasy and alcohol” nights Alcohol Consumption on “No Ecstasy” Nights Compared to “Ecstasy and Alcohol” Nights We compared the “ecstasy and alcohol” nights with nights during which no ecstasy was used (“no ecstasy” nights). “No ecstasy” nights were common, with a total of 613 nights contributed by 134 participants. Across the four timepoints within a night, alcohol consumption peaked at the third timepoint, with an average of 3.54 drinks. The average number of drinks consumed over the course of an entire “no ecstasy” night was 7.0 (SD = 6.6, range 1–56). Figure 5 shows the average number of alcoholic drinks per timepoint for the “ecstasy and alcohol” nights compared with “no ecstasy” nights. Overall, no significant difference was observed in the number of drinks consumed between the two night types (Table 3). However, a significant interaction was found between timepoint and type of night. The increase in alcohol consumption at timepoint 3 and 4 relative to timepoint 1 was larger during “no ecstasy” nights than during “ecstasy and alcohol” nights (timepoint 3: B = 0.69, 95% CI [0.23, 1.15] and timepoint 4: B = 0.61 95% CI [0.11, 1.11]). This means that alcohol consumption increased more steeply as the night progressed when ecstasy was not used, whereas the rise in alcohol consumption was less pronounced on nights when ecstasy was also used. Adjustment for country and location did not change the interpretation of our findings. Estimated marginal mean scores and 95% confidence intervals of alcohol consumption (drinks) in “no ecstasy” and “ecstasy and alcohol” nights at the different absolute timepoints. Table 3.Comparison of alcohol consumption and interaction with time on “no ecstasy” nights1 versus “ecstasy and alcohol” nights​Estimate (95% CI) p valueIntercept2.70​“No ecstasy” night0.06 (−0.22, 0.33)0.68Intercept2.77​“No ecstasy” night−0.25 (−0.62, 0.12)0.19Timepoint 20.33 (−0.05, 0.70)0.09Timepoint 3−0.08 (−0.46, 0.31)0.70Timepoint 4−0.59 (−0.98, −0.20)0.003“No ecstasy” night1 T20.06 (−0.37, 0.49)0.79“No ecstasy” night1 T30.69 (0.23, 1.15)0.004“No ecstasy” night1 T40.61 (0.11, 1.11)0.02 1“No ecstasy” nights may also include drugs other than ecstasy. Comparison of alcohol consumption and interaction with time on “no ecstasy” nights1 versus “ecstasy and alcohol” nights 1“No ecstasy” nights may also include drugs other than ecstasy. Sensitivity analyses showed higher alcohol consumption over time during “alcohol only” nights compared with “ecstasy and alcohol” nights (online suppl. Table 4). When we excluded nights in which other drugs were used, we found a similar (somewhat larger) increase in alcohol consumption over time in the “no ecstasy nights” compared to the “ecstasy and alcohol” nights (online suppl. Table 5). Mood Effects on “No Ecstasy” Nights Compared to “Ecstasy and Alcohol” Nights Figure 6 shows the observed mood effects for “ecstasy and alcohol” nights compared with “no ecstasy” nights. At timepoints 3 and 4 relative to timepoint 1, participants reported feeling more happy and relaxed, and less tired, bored and stressed during “ecstasy and alcohol” nights than during “no ecstasy” nights (Fig. 6; Table 4). There was a significant interaction-effect between timepoint and type of night for both the positive and negative mood scales. At the third and fourth timepoints compared to the first timepoint, participants reported overall higher positive mood and lower negative mood during “ecstasy and alcohol” nights than during “no ecstasy” nights. Adjustment for country and location did not change the interpretation of our findings. Estimated marginal mean scores and 95% confidence intervals of 7-point mood scales in “no ecstasy” versus “ecstasy and alcohol” nights at the different absolute timepoints. Table 4.Comparison of mood and interaction with time on “no ecstasy” nights1 versus “ecstasy and alcohol” nights​Happy p valueRelaxed p valueTired p valueBored p valueStressed p valueestimate (95% CI)estimate (95% CI)estimate (95% CI)estimate (95% CI)estimate (95% CI)Intercept6.17​5.68​2.66​1.53​1.71​“No ecstasy” night−0.42 (−0.50. −0.33)<0.001−0.18 (−0.29. −0.07)0.0011.29 (1.12, 1.46)<0.0010.47 (0.37, 0.57)<0.0010.22 (0.14, 0.30)<0.001Intercept6.00​5.51​2.82​1.72​1.89​“No ecstasy” night−0.25 (−0.41, −0.10)0.0010.04 (−0.16, 0.23)0.700.93 (0.65, 1.22)<0.0010.27 (0.10, 0.44)0.0020.07 (−0.08, 0.21)0.36Timepoint 20.10 (−0.09, 0.29)0.290.03 (−0.21, 0.27)0.81−0.18 (−0.50, 0.14)0.27−0.22 (−0.43, −0.00)0.05−0.08 (−0.26, 0.09)0.35Timepoint 30.28 (0.09, 0.47)0.0050.27 (0.02, 0.52)0.03−0.38 (−0.71, −0.06)0.02−0.31 (−0.53, −0.09)0.006−0.35 (−0.53, −0.17)<0.001Timepoint 40.29 (0.10, 0.49)0.0030.38 (0.13, 0.63)0.0030.00 (−0.32, 0.33)0.98−0.26 (−0.48, −0.04)0.02−0.32 (−0.50, −0.14)<0.001“No ecstasy” night1 T2−0.03 (−0.25, 0.18)0.77−0.04 (−0.32, 0.23)0.750.44 (0.08, 0.81)0.020.21 (−0.04, 0.45)0.100.00 (−0.20, 0.20)0.98“No ecstasy” night1 T3−0.28 (−0.50, −0.05)0.02−0.31 (−0.61, −0.02)0.040.81 (0.42, 1.20)<0.0010.31 (0.04, 0.57)0.020.23 (0.02, 0.45)0.03“No ecstasy” night1 T4−0.56 (−0.81, −0.32)<0.001−0.87 (−1.19, −0.55)<0.0011.02 (0.60, 1.44)<0.0010.39 (0.11, 0.68)0.010.49 (0.26, 0.72)<0.001​positive mood p valuenegative mood p valueestimate (95% CI)estimate (95% CI)Intercept0.28​−0.49​“No ecstasy” night−0.27 (−0.35, −0.19)<0.0010.63 (0.54, 0.71)<0.001Intercept0.14​−0.30​“No ecstasy” night−0.10 (−0.25, 0.04)0.160.41 (0.26, 0.55)<0.001Timepoint 20.04 (−0.13, 0.22)0.63−0.16 (−0.32, 0.01)0.07Timepoint 30.23 (0.05, 0.41)0.01−0.35 (−0.51, −0.18)<0.001Timepoint 40.30 (0.12, 0.48)0.001−0.18 (−0.34, −0.01)0.04“No ecstasy” night1 T2−0.03 (−0.23, 0.17)0.780.19 (0.01, 0.38)0.04“No ecstasy” night1 T3−0.27 (−0.48, −0.05)0.020.43 (0.23, 0.63)<0.001“No ecstasy” night1 T4−0.64 (−0.87, −0.40)<0.0010.57 (0.35, 0.79)<0.001 1“No ecstasy” nights may also include use of drugs other than ecstasy. Comparison of mood and interaction with time on “no ecstasy” nights1 versus “ecstasy and alcohol” nights 1“No ecstasy” nights may also include use of drugs other than ecstasy. Sensitivity analyses showed that when the “no ecstasy” group was split into “other drugs and alcohol” and “alcohol only,” both groups scored lower on most positive mood scales and higher on negative mood scales compared to the “ecstasy and alcohol” group (online suppl. Table 6). Over time, these differences were most pronounced between “alcohol only” and “ecstasy and alcohol” nights. When we excluded nights in which other drugs were used, we found a similar (albeit somewhat larger) decrease in positive mood and increase in negative mood over time in the “no ecstasy nights” compared to the “ecstasy and alcohol” nights (online suppl. Table 7). Discussion This study examined alcohol-ecstasy co-use patterns among British and Dutch frequent electronic dance event visitors by means of EMA. Alcohol-ecstasy co-use was very common, and participants generally drank more before and during ecstasy use compared to after ecstasy use. The significant interaction between timepoint and type of night indicates that drinking trajectories differed depending on whether participants used ecstasy or not. Specifically, alcohol consumption increased more sharply across the night when no ecstasy was taken compared to when ecstasy was used. Although total alcohol intake was higher on nights when ecstasy was used, alcohol consumption within those nights declined after ecstasy intake. Clear mood effects were also found. Overall, positive mood states (happy and relaxed) were higher on ecstasy-alcohol co-use nights compared to nights without ecstasy use. Negative mood states (stress, boredom, and tiredness) showed the opposite pattern. These were lower on ecstasy nights than on non-ecstasy nights. This study uniquely contributes to the literature by presenting the first evidence on both the prevalence of ecstasy-alcohol co-use and the within-night dynamics of this behaviour in a European nightlife setting, offering insights beyond what is known from prior pharmacological research on mood effects and retrospective questionnaire studies. Current findings support results from a previous EMA study that examined co-use patterns in an American study sample [12]. Our findings are also consistent with studies indicating that co-use with alcohol is very common when ecstasy is used [4, 5, 12, 19, 20]. Moreover, our findings are in accordance with research showing that stimulant use is associated with excessive alcohol use [21]. We add to the existing literature by assessing concurrent mood effects, which are particularly relevant given that psychological states associated with drugs and alcohol are important motivational drivers of drug-taking behaviour [22, 23]. Irrespective of ecstasy use, the average number of consumed drinks was higher compared to common cut-off values for binge drinking (>4 drinks of alcohol [for females] or 5 drinks of alcohol [for males] during one session). This insight is of relevance as higher levels of alcohol use likely enhance ecstasy-associated risks such as toxicity and health incidents [7, 24, 25]. The high level of availability of alcohol and the high level of acceptance of drinking in nightlife settings are contributing factors to the high levels of co-use of both substances. It has also been suggested that the consumption of alcohol contributes to the pleasurable and rewarding effects of ecstasy [14, 23]. This is also in line with the results of the current study, which showed higher levels of positive mood effects on ecstasy-alcohol co-use nights compared to nights without ecstasy use. The (expected) additional reward may therefore be an important underlying factor contributing to co-use behaviour. To further interpret the observed interaction between timepoint and type of night, we considered several mechanisms that may explain why the increase in alcohol consumption was less sharp on nights where ecstasy was used, suggesting that ecstasy use moderated the typical within-night escalation in drinking. One possibility is that people intentionally reduce their alcohol use to avoid the dampening effects of alcohol on the stimulating effects of ecstasy. They may also feel less motivated to drink alcohol because the positive mood effects of ecstasy reduce the need to continue drinking alcohol. Another possible mechanism is that people may moderate their drinking to minimise potential harms associated with co-use [26]. A final possible explanation is that alcohol may be partly replaced by the consumption of water. This accords with ecstasy harm-reduction messages advising users to drink sufficient amounts of water, but not excessive amounts that could cause hyponatremia [20, 27]. At the same time, reverse or bidirectional associations are also plausible. For example, nights characterised by steeper increases in alcohol consumption may be less likely to include ecstasy use later in the evening. With respect to mood, we observed that on nights when people consumed ecstasy, they rated themselves as more happy and relaxed, and less tired, bored and stressed than on nights when no ecstasy was used. These positive mood effects are likely the result of the acute pharmacological effects of ecstasy and are in line with findings reported in controlled administration studies [28, 29]. However, given the naturalistic setting of this study, no strong inferences can be made with regard to the time course of the mood effects relative to the moment of intake. It is also not possible to infer to what extent factors other than alcohol and drugs may have influenced mood. On the other hand, the ecological assessment method and the recruitment of participants from two countries may increase the extent to which the findings can be generalised to real-world contexts outside this study. Moreover, these results accord with a wealth of scientific and anecdotal evidence that people feel good while going out and that ecstasy use contributes to this [2, 23, 30]. From a prevention and harm-reduction message perspective, it is important to enhance awareness concerning the high prevalence of combined use. Harm-reduction messages could focus on communicating refraining from alcohol consumption while planning to take ecstasy. However, if people do wish to also consume alcohol, it is advisable to refrain from binge drinking. The present study has several limitations. First, we only investigated ecstasy-alcohol co-use, even though ecstasy is frequently combined with other drugs, as also observed in this study. However, the low frequency of these additional co-use patterns prevented further analyses, and the presence of other drugs limits inferences about pure ecstasy effects or specific alcohol-ecstasy patterns. Second, although the 2-h recall window is shorter than in most substance-use research, it may have obscured shorter-term fluctuations in pharmacological or general mood effects. Relatedly, the 2-h EMA interval and the recoding of assessments near the intake moment into a single category reduced temporal precision and constrained the modelling of immediate within-night mood dynamics. Third, data quality may have been affected by the fact that all participants were under the influence of alcohol and/or other drugs, which may have contributed to missing values. Fourth, there is the absence of systematic measurement of contextual factors such as event size, motivation for going out, and social setting. Because these factors can meaningfully influence both substance use and mood, their omission restricts the interpretability of the results. Fifth, there is a risk of sampling selectivity, meaning that individuals who choose to participate in an EMA study may differ systematically from those who do not. A final limitation is that four consecutive weeks of participation in an EMA study require substantial commitment and effort from participants, which may further contribute to sampling selectivity. At the same time, EMA enhances internal validity by reducing recall bias and capturing processes in real-world settings. In conclusion, the present findings contribute to the evidence base on ecstasy-alcohol co-use patterns. Results suggest that co-use is common, and alcohol consumption levels decrease after the intake of ecstasy. Positive mood states increase over the course of an ecstasy night and negative mood states decrease. These mood effects may act as motivational drivers for use and are therefore relevant for developing harm reduction strategies for people who use ecstasy recreationally. Acknowledgments We like to thank all our participants who took part in the study. Statement of Ethics The study protocol was evaluated and approved by both the Trimbos Ethics Board (Registration No. 201609_41-1619) and the UCL Research Ethics Committee (Registration No. 10437/001). All participants provided written informed consent. Conflict of Interest Statement Dr. Matthijs Blankers was a member of the journal’s Editorial Board at the time of submission. The remaining authors have no conflicts of interest to declare. All financial, personal, or professional relationships that could be perceived to influence the work have been disclosed. Funding Sources This study was funded by ZonMw/ERANID European Research Area Network on Illicit Drugs – Towards Integrated European Research in Illicit Drugs. The UK data collection and delivery of participants’ rewards were funded by the National Institute for Health Research (NIHR) Policy Research Programme (project ref. PR-ST-0416-10003), again within the ERANID network. The funder had no role in the design of the study and during its execution, analyses, interpretation of the data or the decision to submit results. Author Contributions Dr. Desirée B. Spronk contributed to data interpretation and prepared the manuscript. Ruben J.J. van Beek, MSc, contributed to study coordination, data collection, data preparation, and data interpretation. Dr. Robin L.A. Smits contributed to data preparation, data analysis, and prepared the manuscript. Rosa Andree, MSc and Dr. Wouter den Hollander contributed to data preparation, data analyses, and data interpretation. Prof. Tom P. Freeman and Prof. H. Valerie Curran contributed to the study design and supervision of data collection. Dr. Jon Waldron and Dr. Meryem Grabski contributed to study coordination and data collection. Prof. Matthijs Blankers contributed to data analysis, data interpretation, and prepared the manuscript. Dr. Margriet W. van Laar contributed to study design, supervision of data collection, and prepared the manuscript. All authors approved the final manuscript.