An experiment with N = 56 university students investigated the joint effects of manipulated mood state and task difficulty on cardiovascular and electrodermal reactivity during mood inductions and performance on a letter cancellation task. We tested our theory‐based prediction that moods per se do not involve autonomic adjustments whereas mood and task difficulty interact during task performance to determine autonomic reactivity with respect to active coping. Specifically, we anticipated for an easy task weaker reactivity in a positive mood (due to low subjective demand) than in a negative mood (due to high subjective demand). Conversely, we expected, for a difficult task, stronger reactivity in a positive mood (high, but not yet too high, subjective demand) than in a negative mood (too high subjective demand). Adjustments of systolic blood pressure, diastolic blood pressure, and tonic skin conductance level described exactly the predicted pattern. Furthermore, task performance was associated with autonomic reactivity in the difficult conditions.