We tested the hypothesis that juvenile rainbow trout (Oncorhynchus mykiss) would select a temperature colder than their acclimation temperature (16 deg +/-1 deg C) to minimize postexhaustive exercise metabolic demands and enhance oxygen availability. After an initial 3-h exploratory period in a thermal gradient (6 degrees -25 degrees C), fish selected a temperature of approximately 14 degrees C and had a baseline exploratory swimming activity of approximately 60 cm min(-1). Subsequently, experimental (chased) fish were individually removed, exhaustively exercised for 1.5 min, and replaced. Both control (unchased) and experimental fish were allowed to explore the thermal gradient for another 2 h. Immediately after being chased, trout had a metabolic profile that was consistent with being exhausted; levels of plasma and muscle lactate were 4.38+/-0.25 mmol L(-1) and 28.0+/-2.0 mmol kg(-1), respectively, and levels of muscle glycogen, adenosine triphosphate, and phosphocreatine were 3.89+/-0.95, 4.23+/-0.62, and 3.07+/-0.73 mmol kg(-1), respectively. Although exploratory swimming activity of the chased fish was significantly lower (by 81%) as compared with control fish during the first 5 min postchase, differences in the mean, median, and mode values for selected temperatures during the next 2 h were neither large (<1 degrees C) nor significant (P>0.05). Contrary to our initial hypothesis, these findings suggest that juvenile rainbow trout do not select a colder temperature to decrease metabolic rate following exhaustive exercise. Instead, rainbow trout selected a temperature marginally cooler than their acclimation temperature (16 degrees C) regardless of whether they had been previously exhausted.