OBJECTIVES: To determine: (a) whether Scottish drug users on methadone maintenance use heroin less frequently than their peers following other forms of drug treatment; and (b) to what extent those on methadone maintenance 'top up' with heroin. DESIGN: A cohort study followed-up for 33 months from 2001 to 2004. METHODS: Four hundred and ten interviewees who responded at all four interview sweeps, recruited as new treatment entrants from 28 drug treatment agencies across Scotland. RESULTS: Sixty-eight of the 401 interviewees had commenced an episode of methadone-maintenance treatment at the start of the study. There was no significant difference between the methadone-maintained sample and the other interviewees in their propensity to abstain from heroin use, nor was there any difference between the two groups in the mean reduction over time in their self-reported dependence on drugs. However, if the outcome measure used is the change (between baseline and 33 months) in the number of days that the interviewee reported having used heroin in the previous 3 months, the reduction in the number of days that heroin was used was significantly greater (52 days) in the methadone-maintained group than in the rest of the sample (36.4 days). This fall in the number of days of heroin use was greater still if the comparison was restricted to those who had continued on methadone-maintenance treatment, although 67.4% of those still on methadone maintenance had 'topped up' with heroin at some point in the 3 months prior to 33-month follow-up. Those on higher maintenance doses were not significantly more likely to have reduced the number of days on which they used heroin compared with those on lower doses, and those still on methadone maintenance were not more likely to have reduced their criminality (measured by the number of days on which they committed acquisitive crimes in the previous 3 months) compared with the rest of the sample. CONCLUSIONS: Methadone-maintained drug users are not more likely to achieve abstinence than drug users receiving other forms of treatment, but they are significantly more likely to achieve a reduction in the frequency of their illicit drug use; they 'top up' on methadone, but the frequency of their illicit drug use is less than that of drug users in other treatment modalities. These data confirm the value of methadone-maintenance services as part of a 'mixed economy' of services for the treatment of drug use.