PURPOSE: Firstly, to identify whether saliva flow rate, osmolality, and total protein are potential markers of hydration, we compared changes in these parameters with changes in plasma osmolality during progressive dehydration. Secondly, we compared the sensitivity of saliva parameters to track hydration changes with the sensitivity of urine osmolality. Thirdly, to test the hypothesis thatdehydration, rather than neuroendocrine regulation, is responsible for the decrease in saliva flow rate during prolonged exercise, we compared flow rate and catecholamine responses to prolonged exercise with and without fluids. METHODS: colon; Fifteen males (plasma osmolality 289 +/- 4 mOsmol x kg(-1); mean +/- SD) exercised (30 degrees C, 70% RH) with no fluid intake (NFI) until body mass loss (BML) of 1.1, 2.1, and 3.0% and on another occasion with fluid intake (FI) to offset losses. RESULTS: colon; Plasma and urine osmolality increased during NFI (plasma osmolality 3.0% BML: 298 +/- 4 mOsmol x kg(-1); P < 0.01). Saliva flow rate decreased (P < 0.01), saliva total protein increased (P < 0.01), and saliva osmolality increased from preexercise (50 +/- 11 mOsmol x kg(-1)) to 3.0% BML (105 +/- 41 mOsmol x kg(-1)) during NFI (P < 0.01). Saliva osmolality, urine osmolality, and saliva total protein correlated strongly with plasma osmolality during dehydration (r 0.87, 0.83, and 0.91, respectively; P < 0.01). During the FI trial, saliva flow rate and osmolality remained unchanged. Plasma catecholamine concentration increased during exercise (P < 0.01) with no difference between trials. CONCLUSIONS: colon; Saliva osmolality and total protein appear to be as sensitive as urine osmolality to track hydration changes during hypertonic-hypovolemia. These results also suggest that dehydration has a greater involvement in the decrease in saliva flow rate during prolonged exercise than neuroendocrine regulation.