Despite the 2009 decentralization, inter-regional income disparities have remained a contentious issue in Indonesia. Using a newly-constructed regional data set, this study examines whether Indonesian provinces have been converging, which implies that poorer provinces have had higher growth rates than wealthy provinces, and whether decentralization has promoted convergence across regions. The study finds that variations in worker productivity across 26 Indonesian provinces have diminished over time from 0.780 in 1992 to 0.708 in 2000, then declined to 0.666 in 2007, indicating evidence of sigma (s) convergence. Based on absolute (unconditional) beta (ß) convergence hypothesis, regression analyses reveal that the estimated parameters on initial worker productivity for the 1992-2000, 2000-2007 and 1992-2007 periods were -0.0127, -0.0105, –0.0117, respectively. This suggests that absolute beta convergence has also occurred, which means that poor, Indonesian provinces have had higher growth rates than wealthier provinces, thus enabling catch up. However, both sigma convergence and absolute beta convergence measures show that the rate of convergence during pre-decentralization (1992-2000) is higher than that during post-decentralization (2000-2007). Finally, the study applies panel data techniques to estimate the impact of covariates on provincial growth of worker productivity during the 1993-2007 periods. The results show that, first, beta convergence also occurred in Indonesia in conditional sense. Second, foreign direct investments, within-region inequality, trade openness, and oil and gas activities have had positive impact on provincial productivity growth.