Cross-section or short-panel econometric techniques typically used to examine Gibrat’s Law of Proportionate Effect suggest that some degree of mean reversion exists, but may exaggerate the apparent randomness of corporate growth. We argue that a more natural way to explore the long-run distribution of firm sizes is to examine data on the growth of particular firms over long periods of time. Using a sample of 77 UK firms’ real total net assets data observed continually for more than 30 years, our conclusions are that growth rates are highly variable over time, but differences in growth rates between firms do not persist for very long. Further, firms show no tendency to converge to either a common size or to a pattern of stable size differences over time. These results are compared and contrasted with standard approaches using the same data, which suggest firms reach and maintain stable positions in a skewed size distribution.