Continuously recorded amplitude-integrated EEG (aEEG) traces of 77 preterm babies were analysed retrospectively, to study the effect of different sedative drugs over a 24-h period. Thirty-seven babies were treated with phenobarbitone, 18 received morphine and 22 babies received no regular sedation. A "burst" was defined as a discharge of integrated amplitude greater than 10 microV and maximum interburst intervals in 10-min epochs over a 2-h period were measured. Maximum interburst was prolonged in babies given either morphine or phenobarbitone for sedation. Administration of a single dose of diazepam for intubation had a marked additive effect on the EEG depression caused by the base sedative and prolonged the effect for 11 to 12 h after drug administration. We conclude that the effect of sedative drugs must therefore be accounted for when interpreting records of quantified EEG for 12 to 24 h after drug administration.