Empirical studies show low pecuniary returns of switching from wage employment to entrepreneurship. We reconsider the pecuniary gains of this switching by employing a two-stage procedure, where the randomness in the timing of inheritance transfers is used as an exclusion restriction to identify causal effects. The model is estimated on data covering the whole Norwegian population of individuals matched to the entire population of firms established in the period 2002-2011. The results indicate that the average returns to entrepreneurship are significantly negative for individuals entering entrepreneurship through self-employment and modest, but significantly positive, for incorporated startups.