The purpose of this paper is to demonstrate how multiple linear regression provides a viable statistical methodology for dealing with meta-analysis in general, and specifically with the issues of nonindependence and design complexity, such as multiple treatments. Since the F-test and t-test are special cases of the general linear model, approaching meta-analysis from a regression perspective allows the researcher much more flexibility in the questions that can be asked, while insuring that the appropriate error terms and degrees of freedom are used. A number of full and restricted mathematical models are illustrated. A test of effect sizes over and above the dependency of data is presented. (GDC)