Quantitication of stress requires the use of a stress indicator that is easy to measure, and which can be readily interpreted in terms of the potential long-term effects to an organism. This study evaluates the suitability of metabolic rate as an indicator of the stress response in fish. By comparing the metabolic with the cortisol stress response, the most commonly used indicator of stress in fish, it was possible to assess the suitability of metabolic rate as a stress indicator. Changes in metabolic rate were used to predict the long-term effects of transport-related stressors. This study also detennined the baseline metabolic rates of the tish. The standard and the active metabolic rates of juvenile P. cummersonnii were 0.16 ± 0.02 (mean ± S.D, n = 6) mg O₂g⁻¹h⁻¹, and 0.56 ± 0.04 mg O₂g⁻¹h⁻¹, respectively, whereas the routine metabolic rate for the fish was 0.25 ± 0.03 mg O₂g⁻¹h¹. The relationship between metabolic rate and body weight was described by the equation ϺO₂ = 0.64 W⁻°·³⁸. 24-h oxygen consumption measurements showed that juvenile P. commersonnii exhibited diel rhythmicity in oxygen consumption rate, the higher rates occurring at night and the lower rates during the daytime. The higher nocturnal metabolic activity may have been due to increased activity induced by an endogenous rhythm related to feeding. Diel rhythmicity has direct implications for the measurement of baseline metabolic rates since it could result in overestimation or underestimation of these rates. 24-h continuous oxygen consumption measurements enabled the detection of the rhythmicity in oxygen consumption rate, and thereby ensured a greater degree of accuracy in the estimation of these parameters. The metabolic stress response in juvenile P. commersonnii was best described by the equation, y = -0.0013 x² + 0.0364 x ÷ 0.3052, where x = time after application of stressor, and y = oxygen consumption rate. Using the derivative of this equation, the metabolic stress response was estimated to peak approximately 14 min after application of a simulated capture and handling stressor. Oxygen consumption increased by about 300 % as a result of the stress. Approximately 15 min after application of a similar stressor, plasma cortisol levels in stressed fish was 200 % higher than baseline levels. However, cortisol levels in fish sampled 30 min after the disturbance was similar to the baseline cortisol levels, indicating that full recovery had occurred. Although the patterns in the metabolic and cortisol stress responses were similar, metabolic rate could be measured continuously, thereby ensuring accurate interpretation of the data. Furthermore, increases in metabolic rate during the stress response are a culmination of physiological events from the primary to the tertiary levels of biological organization and are, therefore, easier to interpret in terms of long-term effects on the fish. Different transportation procedures elicited variable degrees of stress in juvenile P. commersonnii. The cost of metabolism attributed to the effects of capture and handling was twice as much as that attributed to acute temperature elevation. Acute temperature decrease resulted in a signiticant reduction in the oxygen consumption rate (ANOVA, P < 0.05). Oxygen consumption by the fish was not affected by fish density (ANOVA: F = 2.002, P = 0.5), or by oxygen depletion at dissolved oxygen concentrations above the critical level. Below this level, however, oxygen consumption decreased linearly with further decrease in dissolved oxygen concentration. These results showed that the highest energetic cost to juvenile P. commersonnii was incurred as a result of capture and handling. The results also showed that by subjecting fish to different stressors, it was possible to categorize them according to their relative metabolic costs to the fish. At 25º C, the effective concentration of 2-phenoxyethanol to fully anaesthetize (Stage IV, McFarland 1960) juvenile P. commersonnii was 0.4 ml l⁻¹ and the most appropriate concentration for deep sedation (Stage II, McFarland 1960) of the fish for at least 24 h was 0.2 ml l⁻¹. A maximum of 3 minutes was required by the fish to recover from the effects of the anaesthetic. There was no correlation between fish weight and the rate of induction of anaesthesia (r² = 0.001, p = 0.3). At the peak of the metabolic stress response, oxygen consumption was twice as high in the un-anaesthetized fish compared to the fish anaesthetized after the application of the simulated capture and handling stressor, suggesting that anaesthetization with 2-phenoxyethanol may have reduced the effect of the disturbance on the fish. Similar oxygen consumption rates for the fish anaesthetized prior to capture and the non-stressed fish suggested that the increases in metabolic rate could be linked to the struggling associated with attempts by fish to escape from the perceived stressor. Anaesthetization of juvenile P. commersonnii with 0.3 ml l⁻¹ 2-phenoxyethanol resulted in a more than 200 % increase in plasma cortisol concentration. The elevated levels of plasma cortisol in the anaesthetized fish suggested a manifestation of 2-phenoxyethanol as a stressor. At the time of capture, cortisol levels in fish that were anaesthetized prior to capture were the same as those measured in the disturbed fish at the peak of the stress response (ANOVA, p = 0.95), suggesting that the anaesthetized fish were already experiencing considerable stress at the time they were captured. Undisturbed juvenile P. commersonnii that were anaesthetized for 1 h also had cortisol levels that were five times higher than those measured in undisturbed-unanaesthetized fish, indicating that the duration of exposure to the anaesthetic had a significant effect on plasma cortisol levels. The results presented in this study demonstrate the usefulness of metabolic rate as an indicator of acute stress in fish. This was achieved by comparing the metabolic and the cortisol stress responses. The ease and accuracy with which oxygen consumption of fish could be measured made it possible to measure the stress response more accurately than by plasma cortisol concentration. It was also possible to monitor metabolic rate continuously over a long duration using polarographic oxygen sensors, thus enabling a better evaluation of the stress response. These results, thus, suggest that metabolic rate measurements could be a more practical way to quantify the effects of acute stressors on juvenile fishes. By detailing the profile of the metabolic stress response in P. commersonnii, this study makes a contribution towards understanding the physiological effects of stress in fishes. The study also contributes towards the quantification of baseline metabolic rates of this species under captivity. This study also contributes towards understanding the effects of 2-phenoxyethanol on the stress physiology of fish. By anaesthetizing fish under different conditions of stress, it was possible to evaluate the effect of 2-phenoxyethanol on the metabolic stress response. The ability of 2-phenoxyethanol to reduce physical activity of the fish, and thereby reduce the impact of acute stress on the metabolic stress response, makes it a good agent for the mitigation of stress during the capture and handling of fish. However, the increase in plasma cortisol concentration during prolonged anaesthetization using this drug suggests that the anaesthetic might be a stressor to fish and may, therefore, not be suitable for long-term sedation.