This study is a meta-analysis of overall recruitment rate (ORR) in probability-based online panel research. In this study, we included 23 general population probability-based online panels (out of 28 identified and described) covering 15 countries and provided a comprehensive overview of their methodological approaches to recruitment. We calculated the average ORR and carried out a meta-regression with selected moderators. On average less than 1 out of 5 eligible sampled people registered as panel members, but the data showed a very high heterogeneity of ORRs. We found that recruitment rates were similar in telephone, face-to-face, and mail modes in the United States (U.S.), but the mail mode performed worse than the interviewer-administered recruitment modes in the rest of the world. Also, we identified a small decrease of ORR over time and provided evidence on how ORRs are significantly lower in the U.S. probability-based online panels.