Abstract In this study, the validity of the Gibrat’s Law, i.e. the Law of the Proportionate Effect, is tested using a balanced panel of nearly one million firms in Turkey between 2005–2016. A random parameters model is considered to make use of the heterogeneity of individual firms. Unlike the somehow weak results of the rejecting studies from the literature, Gibrat’s Law is strongly rejected with the random slope and random intercept model utilizing a complete dataset. Employing the distributional properties of firms and modified Gibrat’s Law, which is being used to justify the recent revival of the Law in the literature, did not alter the result.