The influence of reaction time upon stannous (II) chloride as a reductant in the cold-vapor atomic absorption spectrophotometric method was investigated. In this comparative study, virtually all operating conditions, including volumes and chemicals used, were made identical except for the redesigning of the reaction vessels. The use of a reaction mixing time of 2 minutes (TR method), rather than immediate bubbling with aspiration (IL method) of total mercury from undigested biological samples, has several advantages. There is an average 4.5-fold increase in sensitivity of peak height readings with virtually no loss of mercury vapor from the reaction vessel by diffusion. Using the TR method also eliminated interaction of mercury with the matrices investigated, viz., undigested blood, undigested saliva, undigested urine, and digested hair samples. The average recovery for the TR method was superior to the IL method, 99.3 +/- 3.2% vs. 84.7 +/- 3.3%, p less than 0.05. In addition, mercury concentrations determined with the TR method were identical using either peak area or peak height readings. It was also found that concentrations determined using either the formulae given by Magos and Clarkson (4) and Farant et al. (5) or an aqueous calibration curve were equivalent. Only 0.2-1.0 mL of biological samples was needed for each analysis in the study.