We compared two commonly used primary screening methods (FEP and ZnP) for detecting Pb poisoning in children, used according to three major protocols. The results showed that FEP and ZnP values were comparable for only 36% of the children examined; 54% had FEP greater than ZnP and 10% had ZnP greater than FEP. One would identify approximately twice as many children with Pb greater than 300 micrograms/L by an FEP (500 micrograms/L cutoff) as with a ZnP (400 micrograms/L cutoff) screen. With a cutoff of 350, as compared with 500 micrograms/L for ZnP, one would perform 33% additional confirmatory blood Pb determinations with a 3% increased detection of Pb intoxication. If the number of false-negative Pb specimens is to be minimized, the cutoff "action level" for hematofluorometers should be lowered from the currently recommended 500 micrograms/L of whole blood to 350 micrograms/L. Our long experience and that of other laboratories leads us to recommend a revision of Centers for Disease Control risk categories and cutoff values, depending on whether FEP or ZnP is measured in combination with a patient's blood Pb concentration. Our finding of above-normal values for ZnP and FEP with Pb concentrations less than 300 micrograms/L indicates that iron deficiency constitutes an equally important public-health problem for children in New York City.