Aims To measure the interdose milk to plasma ratio (M/P) of R- and S-methadone during multiple dosing in lactating mothers taking medium to high doses of methadone (>40 mg daily), and to assess likely infant exposure.Methods Eight mother/child pairs were studied. initially during their postpartum hospital stay (miniature milk), and where possible again after 15 days (mature milk). The women were on a methadone maintenance programme with daily doses of greater than or equal to 40 mg day(-1). Venous blood was collected at 0, 1, 2, 4, 6, 8, 12, and 24 h and milk was collected from both breasts at 0-4, 4-8, 8-12, 12-16, 16-20, and 20-24 h after dose, R- and S-methadone were quantified by h.p.l.c. The areas under the plasma and milk concentration-time curves (AUC) were estimated and M/P-AUC was calculated. The relative infant dose of both enantiomers was estimated as the product of drug concentration in milk and an average daily milk intake of 0.15 I kg(-1).Results For miniature milk (n=8) the M/P-AUC for R-methadone was 0.68 (95%, CI 0.48, 0.89) and For S-methadone 0.38 (0.26, 0.50). For mature milk (n=2) the M/PAUC, for R-methadone were 0.39 and 0.54 and for S-methadone 0.24 and 0.30, respectively. The estimated doses of R- and S-methadone that would be received by the infant via miniature milk, were 3.5% (2.05, 5.03%) and 2.1% (13, 2.8%), respectively, of the maternal dose (assuming 50% of each enantiomer in the dose). The relative infant dose for R- plus S-methadone together was 2.8% (1.7, 3.9%).Conclusions Breastfeeding during medium to high dose methadone appears to be 'safe' according to conventional criteria because the dosage is <10%. However because the absolute dose received by the infant is dependent on the maternal dose rate, the risk-benefit ratio should be considered for each individual case. The doses of methadone received via milk are unlikely to be sufficient to prevent the neonatal abstinence syndrome.