A serious limitation to the diagnosis of mild organophosphate poisoning and to the preventive screening of organophosphate-exposed workers has been the large interindividual variability in erythrocyte cholinesterase. This makes it necessary to obtain a pre-exposure baseline measurement of enzyme activity as a basis for evaluating subsequent declines. To evaluate a new battery-operated colorimetric erythrocyte cholinesterase kit, 23 workers at a Mexican pesticide formulation plant were examined. All workers had normal cholinesterase, and exposed and unexposed workers were found to have similar mean cholinesterase levels. Although erythrocyte cholinesterase was found to have a coefficient of variation of 12% (similar to that reported in previous studies), hemoglobin-adjusted erythrocyte cholinesterase had a markedly reduced coefficient of variation (7.4%). The 90% confidence interval (24.9-31.7 IU/g hemoglobin) resulted in a lower normal limit that is 78% of the upper limit. Even if a pre-exposure baseline were high normal but unknown at the time of examination, the supervising clinician can be confident that any person with a normal result will be no less than 78% of baseline. The kit is moderately priced, easy to use in the field setting, and the low variability to the assay should allow improvement in diagnosis, screening, and in the epidemiologic evaluation of exposure.