Research on fisheries bycatch and discards frequently involves the assessment of reflex impairment,injury, or blood physiology as means of quantifying vitality and predicting post-release mortality, but exceptionally few studies have used all three metrics concurrently. We conducted an experimental purse seine fishery for Pacific salmon in the Juan de Fuca Strait, with a focus on understanding the relationships between different sublethal indicators and whether mortality could be predicted in coho salmon (Oncorhynchus kisutch) bycatch. We monitored mortality using a ~24-h net pen experiment (N = 118) and acoustic telemetry (N = 50), two approaches commonly used to assess bycatch mortality that have rarely been directly compared. Short-term mortality was 21% in the net pen experiment (~24 h) and estimated at 20% for telemetry-tagged fish (~48–96 h). Mortality was predicted by injury and reflex impairment, but only in the net pen experiment. Higher reflex impairment was mirrored by perturbations to plasma ions and lactate, supporting the notion that reflex impairment can be used as a proxy for departure from physiological homeostasis. Reflex impairment also significantly correlated with injury scores, while injury scores were significantly correlated with plasma ion concentrations. The higher time-specific mortality rate in the net pen and the fact that reflexes and injury corresponded with mortality in that experiment, but not in the telemetry-tagged fish released into the wild could be explained partly by confinement stress. While holding experiments offer the potential to provide insights into the underlying causes of mortality, chronic confinement stress can complicate the interpretation of patterns and ultimately affect mortality rates. Collectively, these results help refine our understanding of the different sublethal metrics used to assess bycatch and the mechanisms that can lead to mortality.