This paper aims to analyze the income dispersion and test both absolute convergence and conditional convergence of income among 26 provinces in Indonesia during 1984- 2008 using static and dynamic panel data approach. Using the σ convergence analysis indicated that income dispersion measured by coefficient variation occurred in 1984-2008 generally experienced fluctuation. Factors influencing income dispersion rate were the impact of the economic crisis, the period of fiscal decentralization in Indonesia, the impact of the Bali bombing, impact of rising fuel prices in October 2005 and the earthquake in Jogjakarta and Central Java. Dynamic panel data estimation with system GMM produced an efficient and consistent estimator to overcome the problems of instrument validity. In addition, it is also dedicated to minimize the risk of bias due to endogeneity problem. There was a strong indication of the existence of absolute convergence and conditional convergence among 26 provinces in Indonesia during 1984-2008. Thus, there was evidence that the economy of poorer provinces tends to grow faster compared to the more prosperous provinces, and this progress meant that there was a tendency to catch up. Based on the system GMM estimation, it is found that the provinces in Java havefaster speed of convergence comparatively to those outside Java.