PIP: This paper reports the results of placebo-controlled double-blind randomized cross-over trial designed to examine the effects of elimination of cow's milk from maternal diet on rates of infantile colic. Subjects were 20 exclusively breast-fed infants, aged 3-18 weeks, with persistent colic. Each mother was given 600 ml drink to be drunk every day; on 6 of the days the drink contained 300 ml of cow's milk and 300 ml of soya milk; on the remaining days it contained 600 ml of soya milk; the 2 drinks were formulated to be undistinguishable; the trial days were grouped into blocks of 2 days, randomly assigned. Mothers were also instructed to record time and duration of any colic, and to record their own daily diet. Avoidance of cow's milk produced no beneficial effects on the incidence of colic in the babies; but when frequency of colic was related to days on which specific foods were eaten by the mothers the rates of colic were found to be significantly higher on days on which mothers reported eating chocolate or fruit. These results do not support the contention that elimination of cow's milk from maternal diet may influence the likelihood of infantile colics; the source of colic, however, cannot be attributed to a single dietary component, but it may involve a combination of foodstuffs.