Agriculture is a major consumer of water, with up to 88% of the total water consumption in summer in irrigated regions, either in France or, for instance, in Australia. Good water management therefore requires an accurate estimation of regional water demand by agriculture, which depends on both soil and weather conditions and on farmers’ practices. We studied the farmers’ practices that influence maize irrigation : sowing and the choice of cultivar in regard to its earliness. Specifically, we aimed to identify geo-referenced indicators that could be used to estimate the spatial and temporal distribution of the various combinations of sowing date, sowing density, sown area and maize earliness. The study was conducted in a 500-km2 irrigated area in south-western France. We first conducted a quantitative analysis of postal survey data to identify environmental factors and farm descriptors that could determine sowing practices and the choice of earliness of cultivar. We then interviewed a group of farmers to find out the main constraints relevant to the sowing date and earliness of cultivar. We identified variables that can be used as indicators of the spatial variability of the studied practices. Our results show that the spatial distribution of sowing date and cultivar earliness over a region can be estimated from climatic descriptors of the area and structural farm characteristics. The first factor allows estimation of tactical variables, the sowing starting date and the cultivar earliness groups, while the second allows estimation of sowing and earliness choice strategies. This is one of the first studies identifying on a regional scale geo-referenced indicators of a crop management system, and the first that provides a conjunctive estimation of sowing and earliness choice practices on a regional scale. This study suggests that for estimating any crop management system, it is helpful to treat strategic and tactical variables separately.