Experiment 1 presented human subjects with 25 shocks of the same (.5, 1.5, or 2.5 mA:between-subjects design) or different (.5, 1.0, 1.5, 2.0, and 2.5 mA:within-subjects design) intensities to test predictions of the dual-process, cortical-model, and adaptation-level theories concerning terminal electrodermal response (EDR) magnitudes in an habituation paradigm. Dual-process theory correctly predicted terminal EDR magnitudes and relative EDR habituation rates. Dual-process theory was further supported in Experiment 3 when EDR magnitude to a standard intensity shock (1.25 mA) decreased with the intensity of a second comparison shock (0, .5, 1.25, or 2.5 mA) only up to the 1.25-mA level, as the common-elements construct of the dual-process theory predicts. Adaptation level incorrectly predicted that standard stimulus EDR magnitude would decrease as comparison intensity, hence adaptation level, increased. Forewarning subjects of each shock intensity increased EDR magnitude in Experiments 2 and 3 contrary to the cortical-model theory's prediction based on subjective stimulus uncertainty.