Markers on chromosome 4q have recently been shown to be associated with insulin resistance in Pima Indians, a population in which insulin resistance precedes and predicts the development of non-insulin-dependent diabetes mellitus (NIDDM). To examine whether genes in this region could play a major role in susceptibility to NIDDM in other populations, we have examined the allele frequencies of a trinucleotide repeat near the fatty acid–binding protein 2 (FABP2) gene on 4q28–31 in three European populations: Finnish, U.K. Caucasian, and Welsh. The U.K. NIDDM population was selected for insulin resistance by studying patients whose obesity-corrected fasting plasma insulin before treatment was above the 98th percentile. Seven alleles were detected. On cross-tabulation analysis, there were no significant associations between allele frequencies and glucose intolerance in any of the populations. Log-linear analysis of the results from all three populations suggested a moderately significant interaction of glucose tolerance status (normal versus diabetic) and the FABP2 allele (partial x2 = 24, df 6, P = 0.027). The parameter describing the interaction of allele A3 and glucose tolerance status was the only such parameter differing significantly from zero (z-score +2.003, P = 0.046). In both the Finnish and U.K. population, the A3 allele was found ∼ twice as frequently in NIDDM than in control subjects (Finnish control subjects, impaired glucose tolerance, and NIDDM: 12.2, 22.4, and 26.6%, respectively; U.K. control subjects and NIDDM: 7.8 and 14.6%, respectively). In the Finnish populations, no associations were found between FABP2 alleles and plasma insulin levels or with homeostatic model assessment (HOMA) estimates of β-cell function and insulin sensitivity. Specific comparison of those subjects with and without the A3 allele revealed no difference in fasting (47.4 [36–61.8] vs. 47.4 [39–57] pM, geometric means and 95% confidence interval, NS) and 2-h (288 [222.6–373.2] vs. 297 [252.6–349.2] pM, NS) insulin or in insulin sensitivity as assessed by HOMA (47.6 [32.3–70.3] vs. 50.5 [39.8–64.2]%, NS). Similarly, in the U.K. NIDDM population, no differences in clinical or metabolic characteristics were found between those with and those without the A3 allele. In summary, in three European Caucasian populations, there is no evidence for a major genetic influence of the FABP2 locus on the development of diabetes. The weak association of the A3 allele with diabetes does not appear to be because of an effect of this allele on insulin resistance.