EN
The paper examines the validity of Gibrat's law - the Law of Proportionate Effects - to Hungarian agriculture. Using a battery of specifications (OLS, FGLS, WLS, two-step Heckmann selection models, and quintile regression) and four size measures (labour, land, capital and total return), the results strongly reject Gibrat's Law for the full sample. Thus they suggest smaller farms grow faster than large. Chow-type tests reject the null of structural break between the evolution of family and corporate farms, suggesting a common growth path.