Evidence for executive deficits among users of MDMA (Ecstasy) John E. Fisk 1 *, Catharine Montgomery 1 , Philip Murphy 2 and Michelle Wareing 2 1 Liverpool John Moores University, UK 2 Edge Hill College of Higher Education, UK Random letter generation and computation span are tasks known to load on executive, prefrontal resources. Previous research suggests that Ecstasy users are impaired on random letter generation. The current study, employing a larger sample (44 current Ecstasy users, and 59 non-Ecstasy users), together with more effective statistical controls for other drug use, failed to replicate previous findings. Ecstasy users were unimpaired on all measures of random generation performance. A significant difference was obtained on the computation span measure, with Ecstasy users scoring significantly lower than non-Ecstasy users. This difference remained statistically significant following control for various indicators of the use of other drugs including cannabis. The results are discussed in terms of the potential effects that Ecstasy might have on different component executive processes. In our previous paper (Wareing, Fisk, & Murphy, 2000) we found evidence for executive deficits among current and previous Ecstasy users. Ten current Ecstasy users (estimate lifetime consumption 1349 tablets), 10 previous users (estimated lifetime consumption 1280 tablets) and 10 drug-naive controls (matched with both groups for age, gender and education) were compared on a range of working memory tests. Executive functioning was measured using the random letter generation task. Random generation is a relatively demanding cognitive process and is known to place a continuous strain on executive resources (see Baddeley, 1996). Participants are asked to produce letters in a random sequence avoiding alphabetical or well-known sequences (e.g. CIA, BBC or FBI). They are also asked to try to produce each letter with the same overall frequency and additionally, in our study, participants were asked to produce only consonants. The task is repeated three times with, respectively, letters produced at one every 4 seconds, one every 2 seconds and one per second. Both user groups performed significantly worse than controls at the 1-second rate of production and produced more vowel intrusions at all three rates of generation (4, 2 and 1 second). Although these results are interesting * Correspondence should be addressed to John E. Fisk, School of Psychology, Liverpool John Moores University, 15–21 Webster Street, Liverpool L3 2ET, UK (e-mail: j.e.fisk@livjm.ac.uk). 457 British Journal of Psychology (2004), 95, 457–466 q 2004 The British Psychological Society www.bps.org.uk