OBJECTIVE: To investigate whether the insulin resistance index (IR) assessed by homeostasis model assessment (HOMA) is associated with the insulin resistance index assessed by euglycemic-hyperinsulinemic clamp (clamp IR) in type 2 diabetic patients who received sulfonylureas (SUs), as well as in those treated by diet alone. RESEARCH DESIGN AND METHODS: Retrospectively, the association between HOMA IR and clamp IR was analyzed in 80 type 2 diabetic subjects (53 subjects treated with SUs and 27 subjects treated with diet alone). The 80 subjects, selected because they had not received insulin therapy, were among 111 diabetic participants in a clamp study for evaluation of insulin resistance from May 1993 to December 1997 in Osaka City University Hospital. RESULTS: The HOMA IR showed a hyperbolic relationship with clamp IR. The log-transformed HOMA IR (all subjects, r = -0.725, P < 0.0001; SU group, r = -0.727, P < 0.0001; diet group, r = -0.747, P < 0.0001) correlated more strongly with clamp IR than did HOMA IR per se (all subjects, r = -0.594, P < 0.0001; SU group, r = -0.640, P < 0.0001; diet group, r = -0.632, P = 0.0004). The univariate regression line between log-transformed HOMA IR and clamp IR in the SU group did not differ from that in the diet group (slope, -6.866 vs. -5.120, P > 0.05; intercept, 6.566 vs. 5.478, P > 0.05). Stepwise multiple regression analyses demonstrated that the log-transformed HOMA IR was the strongest independent contributor to clamp IR (R2 = 0.640, P < 0.0001). CONCLUSIONS: The HOMA IR strongly correlated with the clamp IR in type 2 diabetic patients treated with SUs as well as in those treated with diet alone.