This review used meta-analytic techniques to integrate findings from 30 independent studies that compared programmed instruction to conventional methods of instruction at the secondary level. The meta-analysis demonstrated that programmed instruction resulted in higher achievement when compared to conventional methods of instruction (average ES=.40). No significant correlation was found between class size and effect size (r=.097, p>=.05). The most important aspect of this meta-analysis is that with this virtually zero correlation, it indicates that programmed instruction maintains its effectiveness over conventional methods of instruction regardless of class size. An appendix contains a data-coding form. (Contains 42 references, 1 table, and 5 figures.) (Author/RT)