The effectiveness offamily therapy is assessed in a meta-analysis of 19 studies. At post-treatment, the average patient with family therapy is better offthan 76% ofthe patients with an alternative treatment, a minimal treatment or no treatment. This effect is comparable to that assessed of other psychotherapies at post-treatment. Our data suggest, that the effect of family therapy increases during the first year after treatment, but that the effect may diminish sharply 18 months after the end of therapy. However, more studies with extended follow-ups will be needed before firm conclusions about the long-term effectiveness of family therapy can be advanced.Along with the growth of family therapy in mental health service, data on its effectiveness have also increased. The literature on these studies has been reviewed several times in order to integrate the different results and to estimate the overall effectiveness of family therapy (Gurman et al., 1986). I n 1987, the first quantitative literature review or 'meta-analysis' on family therapy effectiveness was published (Hazelrigg et al., 1987). During the last decade, meta-analysis enjoyed a growing popularity as a review method in several areas ofpsychology as well as in other sciences. It provides a highly quantitative method for integrating research results and yields conclusions that are also more quantitative, and hence more precise and less arbitrary than the conclusions of most traditional narrative review methods (Rosenthal, 1984). Meta-analysis rests on the basic assumption that it is in the same spirit to pool the outcomes of a group of studies as to pool the outcomes of a group of individual subjects in a sample. Instead of studying a sample of individual subjects, one studies a sample of effect studies, in which each study outcome is seen as * Evelyn Markus, M.A., is a fellow-worker in Clinical