In March 1990, 222 organolead manufacturing workers and 62 nonexposed referents were administered a neurobehavioral test battery that included simple visual reaction time (SVRT). SVRT was measured over 44 trials with interstimulus intervals ranging from 1 to 10 sec in a random but fixed order for all study subjects. Different measures of lead exposure and dose (e.g., recent and cumulative exposure based on personal sampling data, exposed/nonexposed status, recent blood lead and zinc protoporphyrin levels, and peak and cumulative urine lead levels) were examined as predictors of several different parameters of SVRT (e.g., mean, median, truncated mean, and standard deviation of SVRT over 44 trials). The association varied, depending on the measures used for SVRT and lead exposure and dose. In linear regression analyses, the strongest and most consistent associations of lead exposure and dose were observed with the standard deviation of SVRT. In assessing the different exposure measures, strong and consistent associations were observed with blood lead levels at the time of SVRT testing, but not with recent or cumulative exposure measures. That is, stronger associations were observed with measures of relatively recent internal dose (i.e., blood lead level) than with cumulative measures (i.e., cumulative exposure). Future studies using SVRT should consider parameters of SVRT that have not been commonly used to date, such as the standard deviation of the SVRT.