Several lines of evidence from family and twin studies have strongly suggested that genetic factors are involved in the development of diabetic microangiopathy [1--3]. Several candidate genes have been investigated to elucidate genetic factor(s) responsible for the vascular complications, but little is known about the genetic basis of these complications [4,5]. Elucidation of the genetic factors predisposing to chronic vascular complications in diabetes mellitus will permit identification of individuals genetically predisposed to the complications and, in turn, will allow us an effective intervention tailored to the specific underlying abnormalities.The renin-angiotensin system regulates the systemic circulation and local haemodynamics, and also regulates cell growth and matrix production via its action on angiotensin II production. Angiotensin I-converting enzyme (ACE; EC 3.4.15.1) is not only a key enzyme in the renin-angiotensin system but also regulates kinin metabolism [6]. The plasma ACE level is under genetic control and is strongly associated with an insertion/deletion (I/D) polymorphism of the ACE gene, defined by the presence or absence of the 287 base-pair Alu -repetitive sequence in intron 16 [7]. Accordingly, the I/D polymorphism has been investigated as a strong candidate marker for genetic predisposition to diabetic vascular complications. Association studies of the ACE genotype with diabetic complications, however, have yielded conflicting results. Diabetologia (1998) 41: 47--53 Meta-analysis of association of insertion/deletion polymorphism of angiotensin I-converting enzyme gene with diabetic nephropathy and retinopathy Summary An insertion/deletion (I/D) polymorphism in the angiotensin-converting enzyme (ACE) gene has repeatedly been shown to be associated with ischaemic heart disease, but the association of this genetic marker with diabetic microangiopathy is controversial. To assess the association of the genotypes with the development of diabetic nephropathy or retinopathy, we performed a meta-analysis of data from the literature, using Mantel-Haenszel method followed by the Breslow-Day test for assessing homogeneity among data. In a total of 4773 diabetic patients from 18 studies with (n = 2495) and without (n = 2278) renal complications, the D allele was significantly associated with diabetic nephropathy (p < 0.0001) in a dominant model (summary odds ratio 1.32, 95 % confidence interval: 1.15 to 1.51). There was no significant evidence against homogeneity of the odds ratios (c 2 = 18.9, 20 df; p = 0.53). The association was significant both in non-insulin-dependent (p < 0.005) and in insulin-dependent diabetes mellitus (p < 0.05). Likewise, in a total of 2010 diabetic patients with (n = 1008) and without (n = 1002) retinopathy, there was no association of the I/D polymorphism with diabetic retinopathy. These data suggest that the ACE I/D polymorphism affects the risk for diabetic nephropathy, but not for diabetic retinopathy. [Diabetologia (1998) 41: 47--53]