A new method for the determination of copper in biological materials by graphite furnace atomic absorption spectroscopy is presented. This new procedure is an extension of the classic method of standard additions, where the analyte concentration is determined in a series of identical samples to which various amounts of metal standard have been added. The concentration of metal in the original sample is determined from an extrapolation of a plot of absorbance versus added analyte. In the new method, the amount of copper is determined by the method of standard additions for different concentrations of the sample under investigation as well. From an extrapolation of the data, the concentration of copper in the absence of interfering matrix is obtained. Studies with fetal bovine serum demonstrate that the new extrapolation technique is precise. Furthermore, considerably more copper is detected than by the classic method of standard additions applied to a nitric acid treated sample. The matrix effects of phosphate, nitrate, albumin, and serum were also examined. Both phosphate and serum, at physiological pH, decrease the detectability of added copper, while nitrate and albumin were without effect. The accuracy of this method has been verified by determining the extinction coefficients of stellacyanin and azurin. The values obtained, 4.33 X lo3 and 3.75 X 10' M-' cm-', respectively, are considerably different from those determined by the method of standard additions on nitric acid digests of these proteins, but were close to values previously reported and determined calorimetrically.