A survey was carried out to identify the current policies of European cooperative groups with respect to interim analyses, stopping rules and data monitoring of ongoing clinical trials. The policies differ widely, from informal interim analyses distributed among all participating investigators, to planned interim analyses carried out by an independent statistician and scrutinized by a data monitoring committee. Different situations clearly call for different policies: for instance, trials of new drugs in AIDS need to be monitored more closely than trials of non-toxic adjuvant therapies for cancer. Likewise, trials with an immediately measurable end-point (such as the large-scale trials in myocardial infarction) need more intensive monitoring than those in which the outcome assessment requires prolonged follow-up. In all cases, however, it seems useful to articulate explicit data monitoring procedures in the trial protocol. In general, an independent data monitoring committee is essential to advise on the desirability to continue accrual into the trial, or to stop it early.