None of the existing reviews or meta-analyses have focused on personalized learning that accommodates learners' interests. To address this issue, we conducted this meta-analysis to examine the effects of personalized learning by interest on self-reports of interest and cognitive load, retention, and transfer, as well as potential moderators of these effects. Based on 26 interest effect sizes (n = 5,335), 8 cognitive load effect sizes (n = 1,228), 46 retention effect sizes (n = 5,991), and 6 transfer effect sizes (n = 375) from 34 publications, our analysis revealed that a) personalized learning by interest had a medium-to-large effect on interest (g = 0.55), a medium-to-large effect on cognitive load (g = 0.54), a medium effect on retention (g = 0.48), and a medium effect on transfer (g = 0.36); b) the effect on interest was moderated by the diagnostic approach, grain size, and the domain, c) the effect on retention varied across learners from different continents, and d) the effect on retention was larger for quasi-experimental studies than experimental studies. Results are discussed in terms of their implications, limitations, and potential to inform future research.