It has previously been postulated that rapid red cell capillary transit through the human lung plays a role in the mechanism of diffusion limitation in some endurance athletes. Methodological limitations currently prevent researchers from directly measuring pulmonary capillary transit times in humans during exercise; however, first pass radionuclide cardiography allows direct measurement of red blood cell (RBC) transit times through the whole lung at various exercise intensities. We examined the relationship between mean whole lung red cell pulmonary transit times (cardiopulmonary transit times or CPTT) and different levels of flow in 88 healthy humans (76 males, 12 females) from several studies (mean age 31 years). The pooled data suggest that the relationship between CPTT and cardiac index (CI), beginning at rest and progressing through to maximum exercise demonstrates that CPTT reaches its minimum value when CI is about 8.1 l m2 min-1 (2.5-3 times the CI value at rest), and does not significantly change with further increases in CI. Cardiopulmonary blood volume (CPBV) index also does not change significantly until CI reaches 2.5 to 3 times the CI value at rest and then increases roughly linearly after that point. Consequently, the systematic increase in CPBV index with increasing pulmonary blood flow between 8.1 and 20 l m2 min-1 displays an adaptive response of the cardiopulmonary system by augmenting CPBV (and perhaps pulmonary capillary blood volume through distension and recruitment) to offset the reduction in CPTT, as no significant difference in mean CPTT is observed between these levels of flow (P > 0.05). Therefore, these data demonstrate that CPBV does not reach maximum capacity during strenuous or maximum exercise. This does not support the principle of quarter-power allometric scaling for flow when explaining modifications during exercise. Therefore, we speculate that the observed relationships between CPTT, CBPV index and flow may prevent mean CPTT (and perhaps mean pulmonary capillary transit times) from decreasing below the threshold time required for oxygenation. Experimental Physiology (2003) 88.2, 191-200.