Measurement of respiratory muscle blood flow (RMBF) in humans has important implications for understanding patterns of blood flow distribution during exercise in healthy individuals and those with chronic disease. Previous studies examining RMBF in humans have required invasive methods on anesthetized subjects. To assess RMBF in awake subjects, we applied an indicator-dilution method using near-infrared spectroscopy (NIRS) and the light-absorbing tracer indocyanine green dye (ICG). NIRS optodes were placed on the left seventh intercostal space at the apposition of the costal diaphragm and on an inactive control muscle (vastus lateralis). The primary respiratory muscles within view of the NIRS optodes include the internal and external intercostals. Intravenous bolus injection of ICG allowed for cardiac output (by the conventional dye-dilution method with arterial sampling), RMBF, and vastus lateralis blood flow to be quantified simultaneously. Esophageal and gastric pressures were also measured to calculate the work of breathing and transdiaphragmatic pressure. Measurements were obtained in five conscious humans during both resting breathing and three separate 5-min bouts of constant isocapnic hyperpnea at 27.1 +/- 3.2, 56.0 +/- 6.1, and 75.9 +/- 5.7% of maximum minute ventilation as determined on a previous maximal exercise test. RMBF progressively increased (9.9 +/- 0.6, 14.8 +/- 2.7, 29.9 +/- 5.8, and 50.1 +/- 12.5 ml 100 ml(-1) min(-1), respectively) with increasing levels of ventilation while blood flow to the inactive control muscle remained constant (10.4 +/- 1.4, 8.7 +/- 0.7, 12.9 +/- 1.7, and 12.2 +/- 1.8 ml 100 ml(-1) min(-1), respectively). As ventilation rose, RMBF was closely and significantly correlated with 1) cardiac output (r = 0.994, P = 0.006), 2) the work of breathing (r = 0.995, P = 0.005), and 3) transdiaphragmatic pressure (r = 0.998, P = 0.002). These data suggest that the NIRS-ICG technique provides a feasible and sensitive index of RMBF at different levels of ventilation in humans.