Although McCarthy and co-workers put a great deal of clinical effort into their study of children less than 24 months of age who had a temperature of 40°C more (Pediatrics 59:663, May 1977), their insufficient statistical analysis brings us no closer to an appropriate outpatient approach to such children. Their comparison of the sensitivity and predictive value of each test in an independent fashion is similar to trying to predict the winner of a long-distance race without knowing who has the fewest laps to go.