This article investigates determinants of growth of milk production in German dairy farms with the use of event history analysis. This methodology enables the analysts to consider time as a proxy for not measurable effects on growth. The likelihood of a farm's moving from a nongrowth episode into a growth episode is estimated and the impact of various covariates on that likelihood is assessed. The analysis is based upon a balanced panel of annual farm accounts from 616 specialized dairy farms in Germany, covering the financial years 1995/1996-2008/2009. The results from event history analyses are presented for low and high growth rates. For both groups, it was found that the probability of entering a growth episode, defined as the event to be analyzed, increases over time, e.g., as a consequence of an increasing need to improve competitiveness. Moreover, several covariates, such as the share of subsidies in returns, farmer's age, and milk price, had a significant impact on growth in a farm's milk production. The analysis revealed, however, that the effect-direction of some explanatory variables differed between the two groups.