The aim of the study was to compare self reported “ecstasy” use with the results of the analysis of hair harvested from the same users. Subjects were recruited by multisite chain-referral sampling within the 1994–95 “dance scene” in Glasgow. One hundred subjects donated hair after completing a lengthy interviewer-administered questionnaire. Overall gross concordance between self reported “ecstasy” use and discovery of MDMA (or related compounds) in analysed hair did not surpass 59%, and no relationship had a Cohen's kappa of more than 0.08. Within the positive concordant dataset (n = 52), scatter was considerable, with no correlation being significant, and none more strongly positive than −0.0518. The results presented here indicate that, as far as MDMA is concerned, if judged by self-report, hair does not reach a level of apparent accuracy that would permit its use as a general population estimator. However, hair testing is probably more reliable than self-report, and its accuracy could be verified independently if large-scale inter- and intra-laboratory comparative research is conducted.