The clinical effectiveness of l-methadone maintenance treatment (LMMT) carried out using d,l-methadone or l-methadone have been compared with ambulatory heroin-dependent subjects. A total of 40 heroin-dependent subjects, previously maintained on l-methadone in Frankfurt am Main, were divided into two groups under randomised double-blind conditions and received either an equivalent dose of l-methadone as d,l-methadone or remained on the previous l-methadone treatment. Requests for a change in the dose of d,l-methadone and l-methadone were recorded, urine samples for determination of illicit drug use were collected and the individual level of opiate craving was determined over a 22-day observation period. There was no significant difference between the two groups in the number requests for a dose change (dose increase <10%). However, there was a significant increase in heroin use in the group which continued to receive l-methadone. Although there was less variability in opiate craving in the group receiving d,l-methadone, the mean intensity of opiate craving did not differ between the two groups. The mean l-methadone dose:l-methadone plasma concentration ratio, an index of the bioavailability of l-methadone in individual subjects, showed no significant change when the treatment was changed to d,l-methadone. The mean d-methadone:l-methadone plasma concentration ratio was 1.17. There was no significant difference between these ratios for day 15 and day 22. The mean l-methadone:EDDP plasma concentration ratio in the l-methadone group was 22.2 and the d,l-methadone:EDDP plasma concentration ratio was 18.4 . The plasma EDDP concentration in the d,l-methadone group increased 3-fold after starting treatment with d, l-methadone. These findings suggest that d,l-methadone can be used in methadone maintenance treatment of heroin-dependent subjects but that further studies are required to evaluate pharmacokinetic interactions between methadone enantiomers.