When studying the factors which influence stress reactivity in within-subject designs, test-retest reproducibility data is needed to estimate power and sample size. We report such data regarding a new experimental stress protocol, based on simultaneous application of the socially evaluated, bilateral feet Cold Pressor Test (CPT) and the Paced Auditory Serial Addition Task (PASAT). Cardiovascular, neuroendocrine, and subjective (affective) stress responses of 32 healthy males were measured twice, at an interval of one week. The novel protocol induced substantial stress reactivity in all parameters at both test and retest. Cardiovascular reactivity remained unchanged, but cortisol and subjective responses were lower at second stress exposure, with high test-retest stability of neuroendocrine (r>.7) and cardiovascular measures (r = .5 to r = .9). PASAT performance improved. Response attenuation suggests habituation-like and/or learning effects. Data provided by our study demonstrate feasibility and power of this stress protocol for investigating changes in stress reactivity in repeated, within-subject designs.