Introduction: In France, the cross-linking of medical and administrative data from different sources—specifically the Program for the Medicalization of Information Systems (PMSI) and the Long-Term Illness (ALD) registry—is one of the methods used to estimate cancer incidence locally in areas without a cancer registry. The objective of this study is to describe the methodology for selecting PMSI hospital stays and ALD beneficiaries and cross-referencing them, and then to compare the resulting figures with incidence data from the cancer registry area. Materials and methods: The study period covered 2006–2008. Data on 21 cancer sites were extracted from the national PMSI, national health insurance databases, and cancer registries. The selection of information from the PMSI and ALD databases, cross-referenced using a common, anonymous individual identifier, was based on algorithms using ICD-10 cancer diagnosis codes. Results: the percentage of PMSI patients matched with ALD beneficiaries ranged from 19% to 60% depending on the cancer site. The number of patients hospitalized or admitted under ALD, considered to have invasive cancer for the first time and counted in the PMSI-ALD cross-referenced database, was higher than the number of incident cases in the registry area. Conclusion: This description is a necessary step toward using the PMSI-ALD cross-referenced database for subnational estimates of cancer incidence. The estimate will be based on applying, to the geographic areas under consideration, the “cancer incidence/PMSI-ALD cross-referenced indicator” ratio calculated for the registry area.