The use of propofol infusions to sedate children in intensive care units has decreased after reports of deaths from myocardial failure. More recently it has been suggested that propofol might have been prematurely condemned. Information about 18 children who had received propofol infusions and suffered serious unwanted effects was used to define their common features. Three of the deaths occurred in one intensive care unit where propofol infusions had been used between 1987 and 1993. During this period 44 children with respiratory tract infections had been admitted to this unit and sedated for at least 48 h. Nine had received long-term (> 48 h), high-dose (> 4 mg.kg-1.h-1) propofol infusions and three had developed progressive myocardial failure and died. There was a significant association between receiving a long-term, high-dose propofol infusion and developing progressive myocardial failure (Fisher's Exact Test, two-tailed hypothesis, P = 0.0128) although a causative relationship could not be proved.