Colombia is the world’s major producer of coca leaves. An important pillar of Colombia’s efforts to curb the production of this illicit crop is to induce farmers to substitute coca for a viable legal alternative. In this paper we identify the extent to which farmers respond to variation in the price of five of the most promising legal alternatives when deciding how much coca to plant – coffee, sugar, palm oil, cocoa, and banana. We do this using a rich, spatially very detailed dataset that contains yearly information on the amount of coca grown in each of over 31,000 villages (veredas) in Colombia over the period 2001-2018. For identification, we exploit exogenous variation in commodity prices, in combination with detailed information about the soil and climatic suitability of each vereda for growing each crop. We find a significant, robust response of coca cultivation to price changes of coffee and banana: when their price goes up, less coca is found in veredas that are better suited to growing these crops. We discuss why we find this effect for these two crops only. Furthermore, we show that our findings are driven by veredas with better (road) access, those closer to army stations, and those without good access to Colombia's waterways that are heavily used by coca(ine) traffickers.