The cellular extrusion of guanosine 3′,5′-cyclic monophosphate (3′,5′-cGMP) is a unidirectional ATP-dependent process that is inhibited by probenecid, a non-selective transport inhibitor of organic anions. In the present study, various cGMP analogues were tested for their ability to inhibit 3′,5′-cGMP efflux and stimulate the cGMP-selective ATPase in human erythrocytes. The difference in uptake of 1 μM [ 3H]3′,5′-cGMP to inside-out vesicles in the presence and absence of 1 mM ATP at 37° was defined as active transport. Two ATP-dependent components were detected for unlabelled 3′,5′-cGMP (0.01–100 μM) with respective K i of 1.3 ± 0.2 and 280 ± 50 μM (mean ± SEM, N = 3). The high-affinity transport was inhibited by the analogues with a typical pattern: Rp-monophosphorothioate guanosine 3′,5′-cyclic monophosphate (Rp-cGMPS) > 3′,5′-cGMP > 2′- O-monobutyryl guanosine 3′,5′-cyclic monophosphate ( O-mb-cGMP) ≈ N 2-monobutyryl guanosine 3′,5′-cyclic monophosphate ( N-mb-cGMP) ≥ N 2,2′- O-dibutyryl guanosine 3′,5′-cyclic monophosphate (Db-cGMP) ≈ 8′-bromo guanosine 3′,5′-cyclic monophosphate (Br-cGMP) ≈ Guanosine 2′,3′-cyclic monophosphate (2′3′-cGMP) > Sp-monophosphorothioate guanosine 3′,5′-cyclic monophosphate (Sp-cGMPS). A concentration-dependent inhibition was found for the low-affinity transport, but no distinct order of potency was identified. Analysis according to Lineweaver-Burk of active [ 3H]3′,5′-cGMP transport (0.2–2 μM) gave a K m value of 1.5 ± 0.1 μM (mean ± SEM, N = 3). The presence of 10 μM cGMP analogues did not change the ordinate intercept, but made the slopes steeper with a typical order: Rp-cGMPS > 3′,5′-cGMP > N-mb-cGMP ≈ O-mb-cGMP ≈ db-cGMP ≈ 8-Br-cGMP > 2′,3′-cGMP > Sp-cGMPS. Only 3′,5′-cGMP and 2′,3′-cGMP were able to activate the cGMP-specific ATPase, 640 ± 200% and 430 ± 160% (mean ± SEM, N = 5) above basal levels, respectively. The present data show that the binding is less selective than ATPase activation of the cellular cGMP transport system.