The use of a Bayesian approach in evaluating data from clinical trials with many treatment centers and from many studies is discussed. The main distinction between a metaanalysis and an analysis of a multicenter trial is that different studies may have very different designs, while the centers in a multicenter trial usually follow the same protocol. In particular, different studies in a metaanalysis may involve different treatment comparisons, while centers within the same trial usually consider the same treatments. The Bayesian statistical approach focuses on the probability distribution of any unknowns given the available information. An advantage of Bayesian methods is that they allow the use of all available information. In the case of a multicenter clinical trial of the effects of a particular drug, Bayesian methods require assessment of the information available before the trial as a probability distribution. A comparison of the Bayesian and frequentist approaches indicates that Bayesian methods have greater flexibility. Results from nine studies of an antidepressant drug illustrate the Bayesian hierarchical approach. Bayesian updating for multiple treatment studies is the same as for a single treatment. Three tables present study data and 15 figures illustrate the analysis. (SLD)