Many of the multicenter trials in perinatal medicine have been negative in that they have shown no benefit for the intervention relative to the control. Although a negative trial can improve patient care by the avoidance of an unnecessary treatment, most trials are designed with the intent of the intervention improving outcomes. We selectively reviewed a number of recent trials in order to identify why the trials were negative. In general, the preliminary information on which the trials were based, which was a small trial or a meta-analysis of multiple small trials, was not robust or predictive. The weak preliminary information together with limited numbers of patients, problematic primary outcomes and a poor understanding of the biology of neonatal diseases has limited the ability to reliably design trials with positive outcomes.