Although a growing number of studies are examining the relationship between Internet use and political participation, varying study characteristics make the overall effect size difficult to estimate. Using a meta-analysis, we estimated the mean effect size and tested whether the effect size was influenced by study characteristics. Data for this meta-analysis were derived from 56 papers reporting 63 independent studies. Results revealed that Internet use had a weak relationship with political participation (rc = .22). Moderator analyses demonstrated that type of Internet use, Internet use measure, Internet use for news, type of political participation, sample origin, and survey year significantly moderated the relationship between Internet use and political participation. For instance, Internet use including news (rc = .27) had a significantly stronger relationship with political participation than did Internet use excluding news (rc = .19). European samples (rc = .27) had the largest mean correlation followed by North American samples (rc = .23) and Asian samples (rc = .18) in decreasing order of strength of relationship. The theoretical and methodological implications of the findings are discussed.