A meta-analysis was conducted on the data obtained from published articles that have used the conditioned place preference (CPP) paradigm to assess the rewarding effects of morphine, heroin, amphetamine and cocaine in rats. Using a histogram analysis of the data, significant dose-effect curves were evident with all of the drugs examined, except for cocaine. Analysis of the data also revealed that several methodological variables moderated the effect size for CPP, at least with some of the drugs examined. In particular, the following methodological variables significantly moderated CPP effect size: strain of rat used; housing condition (single or group cages); type of apparatus (2 or 3 compartments); preconditioning test (present or absent); route of drug administration; intervening saline trials (present or absent); conditioning trial duration; and drug compartment (nonpreferred, counterbalanced or white). No significant effect size differences were evident using sex, number of drug trials, or test duration as moderator variables in the analyses. These meta-analytic results may be useful to investigators for maximizing the effect size of drug-induced CPP.