A method for the determination of different mercury species in whole blood is described. Inorganic mercury (InHg) was determined in 2 ml of standard solutions or blood samples using head space (HS) injection coupled to atomic absorption spectrometry (AAS) after treatment with concentrated sulfuric and tin(II) chloride as a reductant agent in a closed HS vial. After stirring, the InHg was converted to elementary mercury and carried with a nitrogen flow through a quartz cell heated at 200 degrees C and the absorbance signal was evaluated by AAS. For the determination of methylmercury (MeHg), 2 ml of a standard solution or a blood sample were treated with 10 mg of iodoacetic acid and 0.4 ml of concentrated H2SO4. Then, the MeHg species were HS-injected into a gas chromatograph (GC), separated on a semicapillary column (AT-1000) with a flow of helium, then carried to the quartz cell heated at 1000 degrees C and detected by AAS. The high content of salts in blood samples, where sodium chloride is the major component (0.14 mol l-1), affected the gas-liquid distribution coefficient of both mercury species in the HS vial. A linear calibration graph was obtained in the ranges 1-20 and 1-125 micrograms Hg l-1 added as InHg and MeHg, respectively. The detection limits for InHg and MeHg were 0.6 and 0.2 microgram Hg l-1, respectively. The relative standard deviations for eleven independent measurements were 5% for both mercury species. Recovery values ranging from 98 to 106% for InHg and from 95 to 105% for MeHg and from 93 to 95% for ethylmercury (EtHg) were obtained. The accuracy of the proposed method was also established by the analysis of certified whole blood samples for InHg and MeHg. No difference between the sum of these two species determined by our procedure and the recommended total mercury concentrations in the certified samples was observed. Results for the determination of MeHg and InHg in 30 controls and 30 dentists are presented to illustrate the practical utility of the proposed method.