Three procedures for commercial software PeakFit version 4.12 were evaluated comparatively for deconvolution of partially overlapped and fused peaks in chromatogram segments of the petroleum condensate oil sample. Oil sample analysis suffers from a great number of these types of overlapped peaks. It is found that, three procedures were operating on the basis of unification of the value of peak width throughout the chromatogram segment. This finding point has been used as the basis for this research. The obtained results revealed that AutoFit residual procedure operates on the basis of peak width for well resolved real peak. This procedure gives an acceptable number of the peaks and results matching the experimental data. AutoFit deconvolution procedure operates on the basis of a value of peak width for a hypothetical symmetrical ideal peak and gives a large number of positive false peaks. AutoFit second derivative procedure does not have the ability to deconvolute fused overlapped peaks as a result of lack of consistency in the value of peak width for the different chromatogram segments.