INTRODUCTION: Preventable drug-related hospital admissions (pDRAs) occur frequently but with unknown incidence in type 2 diabetes. The objective of this study was to assess the extent to which Danish health care registries can provide data for specific pDRA indicators in diabetes. MATERIALS AND METHODS: Operationalization of indicators was based on a random 10% sample of the Danish population from 2001-2003 with data from The National Patient Registry, The National Health Insurance Service Registry and The Register of Medicinal Product Statistics. Persons with at least one prescription for an oral antidiabetic agent were identified as type 2 diabetes patients. The frequency and the DRG-value of pDRAs were identified for 2002-2003. RESULTS: The 22 defined indicators were operationalized. Indicator validity was reduced due to lack of laboratory data and data on actual medicine use. The patient material consisted of 9,791 persons and 15,645 person years. 1,198 persons were identified with 9,621 pDRAs. Renal problems preceded by no ACE-inhibitor or angiotensin-2-receptor antagonist therapy were frequent (63.85 (CI 59.89-67.81) per 1,000 person-years), as were secondary AMIs preceded by no ASA, beta-blocker or statin therapy (4.92 (CI 3.82-6.02), 9.52 (CI 7.99-11.05) and 12.66(CI 10.89-14.42) per 1,000 person years). The resulting costs of the identified pDRAs were estimated at DKK 55.7 mill. CONCLUSION: Danish health care registries can provide data for explicit pDRA-indicators in diabetes. However, the validity is reduced as laboratory data and data on actual medicine use are not being registered.