We conducted a study to evaluate the use of parotid salivary lead (Pb-saliva) levels as a surrogate of the blood lead (Pb-B) or plasma lead levels (Pb-P) to diagnose lead exposure. The relationship between these biomarkers was assessed in a lead exposed population. Pb-saliva and Pb-P were determined by inductively coupled plasma mass spectrometry, while in whole blood lead was determined by graphite furnace atomic absorption spectrometry. We studied 88 adults (31 men and 57 women) from 18 to 60 years old. Pb-saliva levels varied from 0.05 to 4.4 microg/l, with a mean of 0.85 microg/l. Blood lead levels varied from 32.0 to 428.0 microg/l in men (mean 112.3 microg/l) and from 25.0 to 263.0 microg/l (mean 63.5 microg/l) in women. Corresponding Pb-Ps were 0.02-2.50 microg/l (mean 0.77 microg/l) and 0.03-1.6 microg/l (mean 0.42 microg/l) in men and women, respectively. A weak correlation was found between Log Pb-saliva and Log Pb-B (r=0.277, P<0.008), and between Log Pb-saliva and Log Pb-P (r=0.280, P=0.006). The Pb-saliva/Pb-P ratio ranged from 0.20 to 18.0. Age or gender does not affect Pb-saliva levels or Pb-saliva/Pb-P ratio. Taken together, these results suggest that salivary lead may not be used as a biomarker to diagnose lead exposure nor as a surrogate of plasma lead levels at least for low to moderately lead exposed population.