The purpose of this study was to use meta-analysis to analyze result of 17 studies which investigated the effects of integrated programs, and 11 studies which examined the effects of exercise programs on pain, depression, and disability. The 28 studies analyzed in this work were selected from the following sources. MEDLINE Search, bibliographies of related studies, main academic journals of nursing in Korea, and journals on arthritis issues. For the analysis of the data, homogeniety of effect sizes which were calculated based on data in the 28 studies was tested and its average effect size was computed by using meta analysis software package which was developed by Song(1992, 1998). The results can be summarized as follows : 1) Homogeneity tests were conducted on integrated programs on pain. In the prelimiary homogeneity tests on effect size of all 17 studies, no homogeneity was found. When homogeneity tests on the effect size of the remaining 15 studies were performed, excluding two studies which had extremely larger effect size compared to other studies, the 15 studies were found to be homogeneous(Q=16.38, p=.23). The obtained average effect size, D(Mean Standardized Difference Between Means), was .25. When homogeneity tests on effect size on pain was conducted for the excercise programs, effect size for all nine studies were found to be homogeneous (Q=7.42, p=.49) and the average effect size D=.30. Therefore, Hypothesis 1 was rejected from the results, that an average effect size of the integrated programs on pain was not significantely different from that of the exercise programs on pain. 2) Since only two studies investigated the effect of exercise programs on depression, comparison between the average effect size of integrated programs and that of exercise programs on depression could not be conducted, and hypothesis 2 could not be tested. Thereby, only the average effect size of integrated programs on depression was obtained. Eight studies were tested to be positive on the homogeneity of effect sizes(Q=18.31, p=.02) at $alpha$<.01 and its average effect size was D=.11. 3) For the analysis of integrated program on disability, 13 studies, except for four which had an extremely large effect size compared to the others were found to be homogeneous at $alpha$<.01 (Q=22.30, p=.04) and the average effect size on disability was D=.16. For analysis of the exercize programs on disability, eight studies, except for one which had an extremely large effect size compared to others, were found to be homo geneous(Q=7.87, p=.34) and the average effect size on disability was D=.60. Therefore, Hypothesis 3 was accepted from the results that an average effect size of exercise programs on disability was significantly larger than that of integrated programs on disability.