Drug sales data have increasingly been used for disease surveillance during recent years. Our objective was to assess the value of drug sales data as an operational early detection tool for gastroenteritis epidemics at national and regional level in France. For the period 2008–2013, we compared temporal trends of drug sales for the treatment of gastroenteritis with trends of cases reported by a Sentinel Network of general practitioners. We benchmarked detection models to select the one with the best sensitivity, false alert proportion and timeliness, and developed a prospective framework to assess the operational performance of the system. Drug sales data allowed the detection of seasonal gastrointestinal epidemics occurring in winter with a distinction between prescribed and non-prescribed drugs. Sales of non-prescribed drugs allowed epidemic detection on average 2.25 weeks earlier than Sentinel data. These results confirm the value of drug sales data for real-time monitoring of gastroenteritis epidemic activity.