The purpose of this meta-analysis of single-case studies was to investigate the effects of general characteristics-related and strategic-instruction-related moderators on the use of strategic instruction in improving the ability to solve word problems among students with learning disabilities. For the data analysis, articles published in Korean and English from 1975 to 2018 were reviewed. After applying the inclusion criteria for the meta-analysis, 18 peer-reviewed articles were selected. The selected papers were categorized by the general characteristics-related and strategic-instruction-related moderators, and the GetData Graph Digitizer 2.26 was used to visually extract single case data. Tau-U and BCT values were derived to calculate the intervention effect size of the study, and nonparametric tests were performed using the R program for moderator analysis. First, the overall effect size of strategic instruction was found to be slightly higher than the median. Second, the effect sizes according to the general characteristics-related moderators showed significant differences in subgroups of publication year, age, intervention session, fidelity, and WWC single case research design standards. Finally, in the case of effect sizes according to the strategic-instruction-related moderators, significant intervention effect size was noted between baseline and intervention phases. However, among most strategic-instruction-related moderators, subgroups did not demonstrate statistically significant differences. Based on these results, we discussed the limitations of the current study and made suggestions for future research.