We examine early approaches to the agency problem by asking when captains owned ships in eigheenth and nineteenth century shipping, and how captain-ownership effected voyage success. For each ship we observe the identity of the owners, the identity of the ship captain, the route pursued for each voyage, and various outcome measures for each voyage. We also have information on vessel construction characteristics, the sailing history of each vessel, and the occupations of the owners. In our estimations, we find that captain ownership is more prevalent for vessels where incentives are more likely to be divergent. (We also find evidence that the captain's wealth constraint influences ownership, consistent with models such as Aghion & Tirole 1994.) As for performance outcomes, we find that captain-owned vessels are less likely to sink, less likely to be captured, and more likely to arrive late than their non-captain-owned counterparts.