This dataset stems from a large-scale psychophysical investigation (N = 547) examining the relationship between interoceptive performance and mental health symptoms. It contains behavioral data from cardiac and respiratory interoceptive tasks, including measures of sensitivity, precision, metacognitive bias, and metacognitive efficiency, alongside comprehensive symptom profiling from eight mental health surveys. The data also includes hierarchical Bayesian modeling outputs, exploratory factor analysis results revealing multi-level mental health dimensions, and network analysis estimates. The dataset's purpose is to test prevailing models proposing that altered interoceptive performance is a transdiagnostic mechanism linking brain-body interaction to psychiatric vulnerability, utilizing psychophysically optimized tasks and robust statistical modeling.