Background: During the last two decades, the scale of computer use has been surprisingly increased. In 2000, 80% of workers stated that they use computer in their daily activities. The tendency to use computer is accompanied with healthy risk and the incidence of musculoskeletal disorders between 10 to 60 percent of computer users. Aims to study, risk assessment by RULA and ROSA methods, surveying the correlation of these methods, and evaluating the prediction ability of these methods in the incidence of musculoskeletal disorders among office workers.Methods: This analytic-descriptive study performed at Behbahan University of Medical Sciences and Imam Khomeini port office in 2016. The sample consists of 236 office workers through simple random sampling method. Body map questionnaire as well as RULA and ROSA checklists were used for data collection. Data analysis was carried out by SPSS software, version 22, and statistical tests including one way ANOVA, Pearson, and Chi-square.Results: The results demonstrated that most of the musculoskeletal disorders are related to trunk, neck and back regions by 40.4%, 39.7%, and 35.4% respectively. The furthest risk distribution in ROSA method is at warning level (67.2%) and at high and very high levels in RULA method (62%). Pearson Test showed a positive significant correlation between these methods (P<0.05). Chi-square Test revealed a significant correlation between musculoskeletal disorders in upper and lower limb with RULA method (P<0.05), but there was no significant relationship with ROSA method (p>0.05). One way ANOVA statistical test divulged a significant relationship between risk of musculoskeletal disorders and gender (P<0.05), but did not find a significant relationship with job experience and education (p>0.05).Conclusion: Due of the weak correlation between RULA and ROSA, it can't be surely stated that one may be used as an alternative to the other; however, RULA is better than ROSA to predict the risk of musculoskeletal disorders.