This paper develops and tests a dynamic optimization model of fishermen’s investment behavior in a limited-entry fishery. Because exit from limited-entry fisheries may be irreversible, the fisherman has an incentive to maintain the right to fish (whether by actually fishing or by purchasing an annual license) even when the fishery is not profitable, in the hope that conditions may improve. This incentive provides at least a partial explanation for excess capacity in fishing fleets, one of the most pressing fisheries management issues in limited-entry (and other) fisheries around the world. To assess the ability of simple financial models to explain observed investment behavior, we develop a two-factor (price and catch) real options model of the decision problem faced by an active fisherman who has the option to exit a fishery irrevocably. The immediate reason for adopting a two-factor model is the hope of achieving greater predictive power, since obviously both price and catch are important to fishermen’s decisions. Another advantage to this approach is that it provides a mechanism by which investment behavior can be linked in a real options framework to exogenous factors that affect price and catch separately. For example, international market forces are likely to affect price while having a negligible effect on a local fish stock, while local fish stock dynamics may affect catch directly but have little influence on prices (assuming the demand for a particular fish is relatively elastic). In a comparison of model predictions about fishermen’s exit decisions to 5059 observed decisions in the California salmon fishery in the 1990s, 65% of the model’s predictions are correct, suggesting this approach may be useful in the analysis of fishing fleet dynamics.