Objective: Apo E polymorphism has been shown to affect lipid profiles in non-diabetic and diabetic populations. We evaluated the relationship between Apo E phenotype and fasting lipid plasma levels in type 2 diabetes patients. Methods: Two hundred and ten French type 2 diabetic patients (115 men and 95 women) without any lipid lowering drugs were studied. Fasting lipids were measured by usual methods and Apo E genotype was established for each patient: PCR was followed by digestion of the amplification product with restriction enzymes and separation of the fragments by polyacrylamide gel. Results: Genotypes epsilon 3/epsilon 3, epsilon 2/epsilon 3 and epsilon 3/epsilon 4, epsilon 2/epsilon 2 and epsilon 2/epsilon 4 were found in 68.1%, 14.8%, 15.7%, 1.0% and 0.5%, respectively. No patient had the epsilon 4/epsilon 4 genotype. Lipid plasma levels were compared between E3 group (60) as a reference and E2 (epsilon 2/epsilon 2 or epsilon 2/epsilon 3) or E4 (epsilon 3/epsilon 4 or epsilon 2/epsilon 4). Total cholesterol, LDL cholesterol and Apo B levels were lower in the E2 group. Total cholesterol, LDL cholesterol and Apo B levels were higher in the E4 group. HDL cholesterol levels were increased in the E4 group, as only previously observed in Japanese populations. Conclusion: These results agree with those already reported in diabetic patients of several western European countries. E4 allele carriers have a greater cardio-vascular risk and this could be partially explained by the metabolic variation in lipid metabolism induced by E4 with higher LDL cholesterol and Apo B levels. These results observed in French diabetic subjects without any lipid-lowering drugs may be used as a reference for other studies performed in France.