The effects of global climate change on agriculture will be diverse and complex. Some important qualitative conclusions emerge from the literature. First, it is important to focus on the rate at which climate changes and the capacity of farmers to adjust, rather than on absolute changes in temperature. Second, given that significant warming is inevitable, it is important to focus on the marginal effects of feasible changes in the rate of warming, rather than on the aggregate rate of warming. With a convex damage function, the expected marginal cost of warming may be large even when aggregate damage, given the expected rate of warming, is close to zero. Third, uncertainty is crucial and remains poorly understood. In particular, modelling of low-probability catastrophic outcomes remains very limited. Finally, it seems likely that global climate change will enhance extremes of all kinds.