Abstract This paper tests for beta-convergence and sigma-convergence in the corporate governance models, using a sample of corporate governance ratings for 198 European corporations listed on the FTSE Eurofirst 300 index. A piecewise linear regression is deployed to select a model and the Poisson pseudo-maximum likelihood estimator is also applied to estimate an exponential model. It concludes that there is statistical evidence of beta- and sigma-convergence within countries and the results suggest that institutional differences between countries are statistically relevant.