The objective of this study was to evaluate the relationship of lamp photometrics to differentiation between brake and presence signals. To assess this relationship, signal identification was evaluated as a function of lamp photometrics under simulated dusk/dawn conditions. The following were the main results: (1) Luminous intensity was a better predictor of signal identification than was average luminance. (2) The likelihood of identifying a signal as a brake signal was a monotonic function of lamp intensity. (3) Reaction time was positively related to the degree of subjects' uncertainty (as measured by the relative likelihood of “brake” responses): reaction time was slowest when the likelihood of “brake” or “presence” responses was close to 50%, and it decreased as the likelihood increased or decreased away from 50%. (4) Reaction time in a condition simulating typical U.S. rear-lighting configuration was significantly faster than in a condition simulating typical European configuration. The present results provide support for retaining luminous intensity as the relevant parameter of automobile brake-lighting specifications. Furthermore, these results argue against reducing the current minimum of 80 cd for the brake-lamp luminous intensity.