A new method has been developed at the Pacific Northwest Laboratory to analyze pulse-height energy spectra from whole-body counter in-vivo examinations that use high-resolution Ge detectors. A simple data transformation and smoothing function is used to calculate background and identify photopeaks for isotopic analysis. This technique is beneficial for routine in-vivo measurement programs because it avoids dependence upon complex spectrum deconvolution, stripping, or other least-squares fitting techniques that complicate the assessment of measurement reliability. An in-vivo measurement spectrum is analyzed by first applying the variance stabilizing transformation to the data in each channel, which results in a variable with unit variance. A background spectrum is then determined by smoothing the transformed data. Finally, peaks are identified whenever the difference between the background spectrum and the transformed measurement spectrum exceeds 2.57 standard deviations. This method of spectrum analysis is especially suited to whole-body counting because background spectra are calculated as an integral part of the analysis procedure. This new technique has been applied successfully to routine in-vivo measurements of Pu, Am, and U. The computations for this new pulse height-energy spectrum analysis procedure are easily programmed on a desk-top or analyzer-based computer. The simplicity of the computational technique is also attractive because of the ease with which results can be verified.