The start and end of the urban phase of the Indus civilization (IC; c. 2500 to 1900 BC) are often linked with climate change, specifically regarding trends in the intensity of summer and winter precipitation and its effect on the productivity of local food economies. The Indus Village is a modular agent-based model designed as a heuristic “sandbox” to investigate how IC farmers could cope with diverse and changing environments and how climate change could impact the local and regional food production levels required for maintaining urban centers. The complete model includes dedicated submodels about weather, topography, soil properties, crop dynamics, food storage and exchange, nutrition, demography, and farming decision-making. In this paper, however, we focus on presenting the parts required for generating crop dynamics, including the submodels involved (weather, soil water, land, and crop models) and how they are combined progressively to form two integrated models (land water and land crop models). Furthermore, we describe and discuss the results of six simulation experiments, which highlight the roles of seasonality, topography, and crop diversity in understanding the potential impact of environmental variability, including climate change, in IC food economies. We conclude by discussing a broader consideration of risk and risk mitigation strategies in ancient agriculture and potential implications to the sustainability of the IC urban centres.