The characteristics of spatial distribution and changing trend of regional innovation capability in China are investigated by using Theil index and spatial Gini coefficient, σ-convergence, and β-convergence, respectively. And furthermore, multiple linear regression and spatial autocorrelation are employed to explore the determinants of regional innovation capability. The findings are demonstrated as follows: there exists evident agglomeration of regional innovation activities, especially clustering in the developed coastal region of China; as a whole, regional innovation capability shows a divergent tendency during the period of 1990 to 2007, in spite of there is a convergent trend in the east and the middle parts of China. The results also suggest that regional economic level, industrial R&D intensity, technology absorption capability, and spillovers from public R&D institutions are important determinants of regional innovation capability.