NAD can be used to assess the adequacy of oxygen availability to the respiratory chain. An enzymatic assay was established for NAD in human muscle biopsy samples. It gave reliable, reproducible results. The variation within and between subjects was less than 12%. Muscle NAD and lactate were determined at rest, and after bicycle ergometry work requiring approximately 75 and approximately 100% VO2 max (six subjects, four tests each). A positive (P less than 0.01) linear relationship between resting muscle NAD and percent slow twitch fibers was found, suggesting that fiber types may have different NAD content. Muscle NAD decreased during submaximal and maximal work (P less than 0.05). A large portion (73%) of the NAD reduction could be accounted for by increased muscle water. No relationship could be established between NAD and lactate. The negative linear relationship (P less than 0.01) between the muscle/blood ratio and percent slow twitch fibers is another indication of the fiber having different metabolic responses to the activity.