We determined the response of intracellular cyclic GMP in human arteries and veins and in smooth muscle cells cultured from these vessels to C-type natriuretic peptide in comparison with atrial natriuretic peptide. Repeated-measures analysis of concentration-response curves. Anesthesia research laboratory. Vascular smooth muscle cells from human blood vessels obtained with Institutional Review Board approval and patient consent. Segments of internal mammary artery and saphenous vein were obtained from patients undergoing coronary artery bypass surgery. Smooth muscle cells were cultured from these vessels. Concentration-response curves of intracellular cyclic GMP were determined and analyzed by two-way analysis of variance with repeated measures. In segments of intact saphenous vein, C-type natriuretic peptide was significantly more effective than atrial natriuretic peptide (16-fold increase in cyclic GMP in response to 1 microM of C-type natriuretic peptide vs. six-fold increase in cyclic GMP in response to 1 microM of atrial natriuretic peptide, p < .05). In rings of intact internal mammary artery, 1 microM of atrial natriuretic peptide (26-fold increase in cyclic GMP over basal value) was more effective than 1 microM of c-type natriuretic peptide (three-fold increase in cyclic GMP over basal value, p < .05). In cultured cells from these vessels, the pattern of response to C-type natriuretic peptide and atrial natriuretic peptide was the same as in the intact vessels. These results indicated that human smooth muscle cells in arteries and veins express both forms of natriuretic peptide receptors but that atrial natriuretic peptide acts primarily on the artery and C-type natriuretic peptide acts predominantly on the vein. Increased concentrations of C-type natriuretic peptide could contribute to venous pooling in septic shock.