While the need to reduce the impacts of pesticides on the environment is widely acknowledged, we still know very little about the trade-offs between agricultural productivity, food security and pesticide risk at the paddock level and how it aggregates to regional and national scales. Recent work on the toxicity hazard of agriculture at the national level using USETox revealed an uneven contribution of commodities and growing regions towards the aggregate national toxicity hazard (i.e. the potential to do harm). While this is useful information, the real measure of impact we need to strive for is pesticide risk, which takes into account the local soil and climate conditions (and mitigative practices, if any) to quantify pesticide fate and transport in the environment. Other recent work on the economic and environmental trade-offs of 26 dryland crop rotations in the subtropical grains zone shows how crop rotation design has a significant effect not only over food production and average profitability, but also over pesticide transport parameters like deep drainage, surface runoff, and SOC change. There, pesticide impacts were represented using the number of fallow herbicide sprays as a proxy, but this has limited value to assess pesticide risk because number of sprays or rate of application is poorly correlated to toxicity hazard. Here we aim to assess pesticide risk of dryland crop rotations in the Subtropical Grain Zone of Australia by feeding the results of APSIM simulations to the Pesticide Impact Rating Index (PIRI) model. Our preliminary results highlight the trade-offs between food production and various measures of pesticide risk (risk to mammals, birds, bees, fish, earthworms and pesticide load transported to groundwater), and how these sometimes-competing goals may be balanced.