In this paper, we modelled the Colombian long run per capita economic growth (1925- 2005) using a Markov switching regime model with both fixed (FTP) and time-varying transition probabilities (TVTP) to explain regime changes in the economic growth. We found evidence of non-linearity in the per capita economic growth, and two different levels in the data associated with depression and sustainable growth regimes were identified. In addition, the hypothesis of fixed probabilities is rejected in favor of the time-varying transitional probabilities, meaning that the correct model is the one with endogenous probabilities, when the probability of remaining in the sustainable growth regime increases with a rise in terms of trade, government expenditures and decreases with capital outflows. On the other hand, increases in government expenditures and terms of trade decrease the probability of being in the depression state while an increase in capital outflows raises such probability. Finally, we found that TVTP model gives more information than FTP model because the probabilities have changed significantly during the period under analysis and the explanatory variables are very informative in dating the evolution of the state of the economy, especially those associated with external shocks.