Alarms are ubiquitous in anaesthetic practice, but their net effect on anaesthesiologists' performance and patient safety is debated. In this study, 27 anaesthesiologists performed two simulation sessions in random order; one session was programmed to include an alarm condition, with a standard, frequent, clearly audible alarm sound. During these sessions, adverse events were simulated and anaesthesiologists' response times to these events were recorded. Perceived workload was assessed with the NASA Task Load Index. Response times to adverse events and perceived workload were similar in both groups. Pooled response times to atrial fibrillation and desaturation were fast, with a median (range [IQR]) of 8 (4–14 [1–41]) s and 9 (6–16 [1–44]) s, respectively. Pooled response times to an ST segment elevation on the ECG and an obstructed intravenous line were significantly slower, with median (IQR[range]) times of 34 (21–76[4–300]) s and 227 (95–399 [2–600]) s, respectively (p < 0.001). This study shows that in a simulated anaesthesia environment, response times to adverse events are similar in the absence or presence of an audible alarm, and that response times to various critical events differ.