The current scientific literature provides inadequate evidence to conclude whether or not cumulative exposure to or absorption of lead adversely affects performance in neurobehavioral tests in adults. One of reasons for this controversy is the lack of studies with good cumulative exposure to or dose of lead. The aims of this study are to compare the neurobehavioral test performances of a group of lead-exposed workers and a referent group, and to study the association of the neurobehavioral test performances with concurrent blood lead levels and cumulative blood lead levels. Fifty lead battery workers and 97 non-exposed (referent) workers from a vehicle maintenance workshop were evaluated on their neurobehavioral performance using the World Health Organization Neurobehavioral Core Test Battery (WHO-NCTB). The geometric mean concurrent blood lead (ConPb) of the exposed and referent groups were 37.1 (range 13.2-64.6) microg/100 ml and 6.1 (range 2.4-12.4) microg/100 ml, respectively. Cumulative blood lead (CumPb) was defined as area under the curve for the number of years each worker was exposed to lead (three workers previous blood lead results were not available). ConPb and CumPb were used to study the association with the neurobehavioral test results. The exposed group had significantly poorer manual dexterity, perceptual-motor speed, and motor steadiness compared with the referents. The standardized partial regression coefficients were higher for CumPb than ConPb for most of the neurobehavioral test results. In the group >35 years old, there were significantly stronger associations between CumPb and Digit Symbol and Trail Making Part A results than for ConPb which are tests of perceptual and motor skills. CumPb was a better predictor than ConPb of the effects of lead on neurobehavioral performances.