For many low-income countries, the impact of structural reforms on economic growth and poverty alleviation crucially depends on the response of aggregate agricultural supply to changing incentives. Despite its policy relevance, the size of this parameter is still largely unknown. This paper discusses the different approaches which may be employed to quantify the agricultural supply response. It turns out that none of these approaches is likely to deliver unbiased estimates. While in cross-country regressions the problem of unobserved country characteristics cannot be fully eliminated, time-series estimations tend to suffer from the Lucas critique. Any comprehensive empirical analysis should thus rely on more than one technique in order to check the robustness of results.