Using a unique dataset comprising information for (up to) 153 firms in the machine building sector in Belarus, we investigate the determinants of firm growth for an economy where state ownership of enterprises is widespread. We use panel data models based on generalizations of Gibrat's law, total factor productivity estimates and matching methods to assess the differences in firm growth between private and state-controlled firms. Our results indicate that labor hoarding and soft budget constraints play a particularly important role in explaining differences in performance between these two groups of firms.