INTRODUCTION: Ecological momentary assessment (EMA) studies show that stressors trigger the onset or increase of psychotic symptoms. These studies, however, predominantly rely on large sampling intervals and self-report assessment. This study aims to identify the autonomic stress-response to psychosis-spectrum experiences by using a one-day high-resolution EMA with continuous skin conductance and heart rate monitoring in a sample with attenuated positive symptoms. METHODS: Sixty-two participants were equipped with a smartphone and wearable sensors to monitor skin conductance level (SCL) and heart rate variability (HRV) for 24 h. Every 20 min, participants answered questions on current stress, hallucination spectrum experiences (HSE), and paranoia. Sampling intervals were categorized into no event, pre-onset, event, pre-offset, and post-offset phases. We contrasted stress, SCL, and HRV between phases using multilevel regression models of sampling intervals nested in participants. RESULTS: For paranoia, we found alterations in the autonomic and self-reported stress response prior to the onset that persisted until the episodes had ended. For HSE, we found no effects. Exploratory separate analyses of the different items aggregated into HSE yielded diverging results for intrusive thoughts, perceptual sensitivity, and hallucinations. CONCLUSION: Physiological parameters are sensitive indicators of the onset of paranoia, which holds implications for preventive mobile interventions. To further explore the autonomic stress-response associated with HSE, further studies of the different HSE are needed.