We address the issue of whether financial structure influences economic growth. Three competing views of financial structure exist in the literature: the bank-based, the market-based and the financial services view. Recent empirical studies examine their relevance by utilizing panel and cross-section approaches. This paper, for the first time ever, utilizes time series data and methods, along with the Dynamic Heterogeneous Panel approach, on developing countries. We find significant cross-country heterogeneity in the dynamics of financial structure and economic growth, and conclude that it is invalid to pool data across our sample countries. We find significant effects of financial structure on real per capita output, which is in sharp contrast to some recent findings. Panel estimates, in most cases, do not correspond to country-specific estimates, and hence may proffer incorrect inferences for several countries of the panel.