The acceptance by children of two pharmacologically identical phenoxymethyl-penicillin potassium and erythromycin ethylsuccinate mixtures, differing only in their flavouring agents, was compared in a randomised blind manner. The time a nurse required to give the drug to a child was recorded and a score of the acceptance was given by the nurse. The mean time to give erythromycin Brand 1 was 29 s (SD +/- 21, number of children = 20) and mean time to give Brand 2 it was 16 s (SD +/- 13, n = 18); the difference was significant (p less than 0.05). The mean scores were 2.9 (SD +/- 0.6) and 3.4 (SD +/- 0.5), difference -0.5 (p less than 0.025). The difference between the two penicillin products was not significant. The time required to give a medicine was a useful way to compare the effect of different flavouring mixtures on the acceptance of drugs by infants and children.