The objective of this research is to estimate the true parameter of gamma distribution by using Maximum a Posterior approach, Bayes approach, Maximum Likelihood approach which are Maximum a Posterior estimator, Bayes estimator and Maximum Likelihood estimator, respectively. These estimators are compared with Mean Square Error for estimation of the true parameter when the data drown from a gamma distribution. In the case of the small sample size, the Maximum a Posterior estimator is quite well as compared with Bayes estimator and Maximum Likelihood estimator base on Mean Square Error. In another case of the sample size appear that Maximum a Posterior estimator and Bayes estimators are less Mean Square Error than Maximum Likelihood estimators. Our results suggest that Maximum a Posterior estimator for estimation of the true parameter of gamma distribution because Mean Square Error of Maximum a Posterior estimator is quite well when it is compared with Bayes estimator and Maximum Likelihood estimator.