Markets are the bedrock of economic theory. The idea that prosperity is formed through the exchange of goods and services is as old as the study of economics itself (Smith 2008). Yet for agrarian societies, market-driven prosperity is far from certain. Economic actors in agricultural food systems continuously face uncertainty and risk -market price risk, climate risk, and political risk to name a few sources. Even if actors account for these risks ex ante, unexpected events (shocks) lie in wait to derail progress towards prosperity. This thesis investigates economic actors' adaptation to risk, responses to shocks, and markets' role in actors' decisionmaking under risk.The thesis presents economic theory and empirical evidence from three countries in the Global South -Tanzania, Peru, and Ethiopia -across different levels of economic actors. First, the thesis starts with understanding how households use production diversification to manage risks ex ante and what the implications of diversification strategies and market access are for household nutrition. Next, the thesis aims to understand intra-household decision-making in agricultural investment under production and market risk and what the implications of gendered decision-making are for family labor requirements and productivity. The thesis then moves to the analysis of ex post responses to risk. It explores how coffee-farming households affected by negative production shocks change their marketing strategies as a coping mechanism. Finally, the thesis zooms out to the market-level to show how market price fluctuations in food markets can be a source of uncertainty and drive regional and national instability and conflict through price effects on producers and consumers.Each chapter presents a stand-alone contribution to distinct strands of economic literature, but together the chapters underscore economic actors' heterogeneity in risk management and responses to shocks. Economic actors are diverse and operate in a myriad of contexts with distinct opportunities and constraints. Diverse contexts and actors necessitate diverse behavior. Despite this feature of the real-world, 'It is odd that heterogeneity does not play a greater role in economic models' (Kirman, 2006). Much of economic theory has been built on the notion of representative actors (Hartley, 1996; Marshall, 1920), and empirical work is often implicitly or explicitly concerned with mean impacts (Carneiro et al., 2002). In recent years, theoretical considerations have been evolving to consider actors' heterogeneity, especially when aggregating microeconomic actors into macroeconomic models (Cristelli, Pietronero, and Chapter