This study was to analyze the effectiveness and the fidelity of intervention programs for elementary students with math learning disabilities or math underachievements. To meet the research purpose, the research questions were established as follows: First, how did the selected studies consist their content elements of the program and how was each study’s fidelity turned out? Second, how were the selected studies’ effect sizes turned out? Third, is there any correlation between the program fidelities and effect sizes of the selected studies in relation to research design? The researcher selected 25 studies on math intervention for elementary students with math learning disabilities or math underachievements. The rubric method was hired to analyze the fidelity of each intervention in order to evaluate the selected studies relatively and qualitatively. The results were as follows. First, the selected studies’ contents fidelities appeared from 62.15 to 80.13 percent point. The average fidelity of studies employing multiple baseline design was 80.13 percent point which was the highest among three research design groups. The research design group which hired a simlpe AB design showed 62.15 percent point. Second, the effect sizes were turned out from 2.78 to 14.84. The research design group which hired a simlpe AB design showed 2.78. The research design group which hired multiple baseline design showed 14.84 as average effect size. Third, the study employing multiple baseline design that was scored highly on the fidelity rubric showed higher effect size. Regrading these findings, there were discussed some aspects and each study’s result in terms of their fidelity and effect size in the discussion.