The "gold standard" measure of the human startle eyeblink response is the ocular electromyogram (EMG). However, EMG measurement is not always feasible, as with special populations or during functional neuroimaging. We evaluated an alternative, nonelectrical, noncontact measure that uses infrared (IR) light reflected from the eye. By simultaneously recording IR and EMG during an acoustic prepulse inhibition of startle paradigm, we were able to directly compare the two measures and evaluate the relative reliability and validity of the IR measure as an index of startle response modulation. Although fewer responses were detected using IR than EMG, both measures were equally sensitive to prepulse modulation of response amplitude, latency, and probability. We conclude that when the goal is simply to assess the effects of a prepulse on the startle response, IR reflectance is an adequate alternative to EMG.