This thesis was aimed to know the difference of students’ learning outcomes using remedial teaching program between peer tutor method and convensional method in MIN Sleman Yogyakarta. This research methodology was quasi experimental research with nonequivalent control group design where the subjects were the fifth grade students who had not passed the passing grade fixing by the school, that is 75 and also obligated to follow remedial program. The students who followed remedial teaching program at MIN Tempel consisted of 18 students from the total 94 students or 19.1%. However, at MIN Yogyakarta 1 consisted of 13 students from the total 33 students or 39.3%. The examining was done using nonpharametric with Man-Whitney test or U test. The result of this research showed that, pair tutor method had more influence than convensional method in increasing students’ learning outcomes. It was proved by the average (mean) of posttest scores in experimental class, 8.11 higher than the average (mean) of posttest scores in control class, 6 at MIN Tempel. Meanwhile, at MIN Yogyakarta 1, the everage (mean) of posttest scores in experimental class was 9.5 higher than the average (mean) of posttest scores in control class, 6. Then, from the analysis result using Mann-Whitney test (U Test) using SPSS 16.0 for windows, it could be seen that test statistics table showed asymp.sig.(2tailed) as high as 0.017 < 0.05, so that Ha was accepted at MIN Tempel. Moreover, at MIN Yogyakarta 1, it showed that asymp. Sig. (2tailed) as high as 0.002 < 0.05, so that Ha wa accepted. It meant that the treatment using pair tutor had more influence in increasing students’ learning outcomes in remedial teaching program.