Paired capillary-venous samples were obtained from 255 women undergoing a glucose challenge test and 116 women undergoing an oral glucose tolerance test. The capillary equivalents for the venous threshold values were calculated by regression analysis. The glucose challenge test predictions of either normal or abnormal agreed in 82%. The sensitivity, specificity, and positive and negative predictive values for the capillary oral glucose tolerance test were 89%, 90%, 62%, and 98%, respectively. These capillary equivalents were then applied prospectively to 147 women undergoing a glucose challenge test and 141 women undergoing an oral glucose tolerance test. The concurrence rate of the glucose challenge test in the prospective group was 90%. The sensitivity, specificity, and positive and negative predictive values for the capillary oral glucose tolerance test were 64%, 95%, 75%, and 92%. When the venous threshold recommendations of the American Diabetes Association were used instead of those standard at our institution, these values increased to 75%, 98%, 83%, and 96%, respectively. The recommended capillary values of the American Diabetes Association were 100% sensitive but had a positive predictive value of only 20%. Based on the prospective group, the cost per case of gestational diabetes identified would decline 63% if both a capillary glucose challenge test and an oral glucose tolerance test were used and 25% if the capillary glucose challenge test and venous oral glucose tolerance test were used. Combining the data set for new regression equations, the following venous-capillary threshold sets emerged: glucose challenge test, 140 mg/dl/150 mg/dl; fasting oral glucose tolerance test, 105 mg/dl/114 mg/dl; 1 hour, 190 mg/dl/211 mg/dl; 2 hours, 165 mg/dl/183 mg/dl; 3 hours, 145 mg/dl/157 mg/dl. The sensitivity, specificity, and negative predictive values for the capillary oral glucose tolerance test with these thresholds were 80%, 97%, 80%, and 97%. In conclusion, capillary glucose testing for diabetes during pregnancy is feasible and cost-effective.