The maximum rate at which fish can take up oxygen from their environment to fuel aerobic metabolism is an important feature of their physiology and ecology. Methods to quantify maximum oxygen uptake rate ((M)over dotO(2)), therefore, should reliably and reproducibly estimate the highest possible (M)over dotO(2) by an individual or species under a given set of conditions (peak (M)over dotO(2)). This study determined peak (M)over dotO(2) and its repeatability in Gulf killifish, Fundulus grandis, subjected to three methods to elevate metabolism: swimming at increasing water speeds, during recovery after an exhaustive chase, and after ingestion of a large meal. Estimates of peak (M)over dotO(2) during swimming and after an exhaustive chase were repeatable across two trials, whereas peak (M)over dotO(2) after feeding was not. Peak (M)over dotO(2) determined by the three methods was significantly different from one another, being highest during swimming, lowest after an exhaustive chase, and intermediate after feeding. In addition, peak (M)over dotO(2) during recovery from an exhaustive chase depended on the length of time of recovery: in nearly 60% of the trials, values within the first hour of the chase were lower than those measured later. A novel and important finding was that an individual's peak (M)over dotO(2) was not repeatable when compared across methods. Therefore, the peak (M)over dotO(2) estimated for a group of fish, as well as the ranking of individual (M)over dotO(2) within that group, depends on the method used to elevate aerobic metabolism.