The paper investigates the decisions of companies to enter new submarkets, analyzing the relationship between submarket concentration and entry through the explanatory power of submarkets. The relationship is analyzed both at the aggregate level for all countries and at the country-specific level using a pharmaceutical company dataset gathered in seven countries in the 1987-1998 period. A Bayesian panel probit model enables us to measure entry into a new submarket using the lagged dependent variable to incorporate potential state persistence. The results obtained at the submarket level indicate a negative correlation between concentration and entry by diversification, although previous literature at market level has indicated conflicting results for the same relationship.