Study effect meta-analysis was used to synthesize the results of explicit instruction on critical thinking (CT). The approach involved the collection of 250 studies (books, articles, dissertations, and abstracts) from the ERIC database and "Dissertation Abstracts International", coding of the study features, calculation of effect sizes, and examination of relations between study features and effect sizes. To be included in the meta-analysis, studies had to compare the performance of two groups of students on some measure of CT. One group received explicit instruction to improve CT, while the other group did not. Each study was coded for its characteristics within three categories: publication history, methodological and setting features, and instructional features. Instructional variables were duration of instruction, subject matter of the course in which the instruction was given, and presentation of instruction via the teacher or some other medium. Programs that focused continuously on thinking skills were distinguished from those that addressed such skills periodically. Programs that taught CT in a subject matter domain were distinguished from those that taught CT generally. Finally, the content of programs was coded as addressed to either internal or external consistency. A total of 20 studies, 19 of which were doctoral dissertations, were found to be suitable for the meta-analysis. Results consistently favored programs that used explicit instruction methods. Intensive programs proved more effective than did programs providing only periodic training in CT. Length of treatment was unrelated to effectiveness. When the content of the CT instruction was internal validity, the average effect size was significantly lower than when both internal and external validity were emphasized. (TJH)