In a placebo-controlled, cross-over study, thirty-seven puerperal women with inadequate production of breast-milk were treated with 5, 10, or 15 mg of metoclopramide three times a day for 2 weeks. Doses of 10 or 15 mg significantly raised maternal serum prolactin; they also increased breast-milk secretion by 42.5 +/- 34.7 (SD) ml and 50.0 + 35.9 ml per feed, respectively. This effect was unrelated to the phase of the puerperium during which treatment was started. The increase in milk secretion was associated with a decreased need for supplementary feeds, and 33% of the infants of these mothers needed no supplementary feeds during treatment. 5 mg doses did not stimulate prolactin milk secretion. Although placebo had no objective effect on the milk yield, 24% of the women judged its effect to be good, and 89% of women on metoclopramide reported a good effect. Seven women on metoclopramide and three women on the placebo complained of slight side-effects. No adverse effects upon the infants were observed. Metoclopramide therapy may be useful for improving poor lactation.