Canada is one of the few countries for which data exist on individual family farms over time. Using these data, researchers have been able to show that much of the microdynamics of family farms in Canada (e.g. changes in size distribution) can be attributed to farmer entry and exit. However, the behavioral aspects of the exit decision received little attention in the literature. A comparison of Canadian and Israeli data could help us understand these behavioral aspects because of the vast institutional differences between the farm sectors in the two countries, and the possible effect of the institutional setup on exit decisions. Comparison of exit patterns may enable us to identify the dependence of farmers' mobility on the institutional setup. This may also have policy relevance. In both Canada and Israel, exit probability decreases with the extent of off-farm work. We conclude that off-farm work is complementary rather than a substitute for farm work, perhaps due to its less volatile nature. Both, Canadian and Israeli farmers over a certain age are more likely to exit as they become older, which is a natural result, but exit probability in Canada rises much faster with age than in Israel. The major difference between Canadian and Israeli farm-exit patterns lies in the farm size. Farm size decreases exit probability in Canada but increases it in Israel. Perhaps this is because Israeli farm exits are less planned in advance than Canadian exits. Institutional constraints on land transactions in Israel may also play a role. © 1999 Elsevier Science B.V. All rights reserved.