Several methods that are promising for studying the neurophysiology of fear conditioning (e.g., EEG, MEG) require a high number of trials to achieve an adequate signal‐to‐noise ratio. While electric shock and white noise burst are among the most commonly used unconditioned stimuli (US) in conventional fear conditioning studies with few trials, it is unknown whether these stimuli are equally well suited for paradigms with many trials. Here, N = 32 participants underwent a 260‐trial differential fear conditioning and extinction paradigm with a 240‐trial recall test 24 h later and neutral faces as conditioned stimuli. In a between‐subjects design, either white noise bursts (n = 16) or electric shocks (n = 16) served as US, and intensities were determined using the most common procedure for each US (i.e., a fixed 95 dB noise burst and a work‐up procedure for electric shocks, respectively). In addition to differing US types, groups also differed in closely linked US‐associated characteristics (e.g., calibration methods, stimulus intensities, timing). Subjective ratings (arousal/valence), skin conductance, and evoked heart period changes (i.e., fear bradycardia) indicated more reliable, extinction‐resistant, and stable conditioning in the white noise burst versus electric shock group. In fear conditioning experiments where many trials are presented, white noise burst should serve as US.