1. The excretion of dothiepin, nordothiepin, dothiepin-S-oxide and nordothiepin-S-oxide into breast milk was studied in eight women. Exposure to drug was measured in five of their infants, and possible drug-related effects were assessed in all eight infants. 2. Using pre-feed milk samples mean (+/- s.e. mean) milk:plasma (M:P) ratios were 0.78 +/- 0.12, 0.85 +/- 0.16, 1.18 +/- 0.29 and 1.86 +/- 0.29 for dothiepin, nordothiepin, dothiepin-S-oxide and nordothiepin-S-oxide, respectively. In post-feed milk samples, the mean M:P ratio for dothiepin (1.59 +/- 0.32) was significantly greater (P less than 0.05) but M:P ratios for the metabolites were similar. 3. Mean total calculated infant daily doses, (in dothiepin equivalents and as a percent of the maternal dose) were 0.58% for dothiepin, 0.23% for nordothiepin, 2.47% for dothiepin-S-oxide, and 1.17% for nordothiepin-S-oxide. 4. Plasma samples were obtained from five infants. In one, both dothiepin and nordothiepin were below their minimum quantifiable levels (2 micrograms l-1) while in four others both dothiepin-S-oxide and nordothiepin-S-oxide were below their minimum quantifiable levels (10 micrograms l-1). No adverse effects were found in any of the eight infants. 5. Use of dothiepin by depressed mothers is unlikely to be a significant hazard to their breast-feeding infants.