A method for obtaining accurate, quantitative 57Fe Mössbauer spectra from biological samples is illustrated stepwise in a data reduction procedure. Exact criteria are presented for deciding when it is necessary to account for the effects of the Beer-Lambert law in the Mössbauer spectra from biological samples. This procedure makes extensive use of the fast Fourier transform and other computer techniques in its data reduction and its curve-fitting stages. A method for optimizing sample thickness is presented. The choice of truncation in Fourier space as a means to numerically stabilize the deconvolution procedure is defended. Several advantages for curve fitting in Fourier space are shown. Maximization of information content is discussed for Mossbauer spectral simulation techniques.