This paper presents a tractable model of firm growth driven by firm-specific productivity shocks, selection of successful firms, and imitation by new entrants. The model generates balanced growth and aligns with the observed size distribution of firms, which follows a Pareto distribution with a tail index around 1.06. The model explains Zipf's law as indicating high entry costs or difficulty in imitation. Calibration based on U.S. data suggests that about half of output growth is due to selection, but the variance of firm growth rates is unexpectedly high. The model also predicts that lower entry costs increase the variety of goods produced but reduce the level of output due to more inefficient firms entering and surviving. The mechanism of imitation ensures that the tail index of the size distribution converges to 1 as entry costs become large, reflecting the stationary nature of the size distribution.This paper presents a tractable model of firm growth driven by firm-specific productivity shocks, selection of successful firms, and imitation by new entrants. The model generates balanced growth and aligns with the observed size distribution of firms, which follows a Pareto distribution with a tail index around 1.06. The model explains Zipf's law as indicating high entry costs or difficulty in imitation. Calibration based on U.S. data suggests that about half of output growth is due to selection, but the variance of firm growth rates is unexpectedly high. The model also predicts that lower entry costs increase the variety of goods produced but reduce the level of output due to more inefficient firms entering and surviving. The mechanism of imitation ensures that the tail index of the size distribution converges to 1 as entry costs become large, reflecting the stationary nature of the size distribution.