The effectiveness of different statistical distribution functions (Gaussian, log-normal, gamma and Weibull) is examined for the deconvolution of chromatograms with overlapping peaks generated under diverse instrumental set-ups and operating conditions. Some other modified Gaussian functions (Littlewood and Grubner) are considered for comparison. Standard mixtures of phenanthrene and anthracene were analysed by gas and liquid chromatography (GC and LC) and by mass spectrometry coupled to GC and LC. Several instruments including columns of different dimensions and polarity and diverse types of injectors and detectors were used to produce the data set. In general terms, the log-normal function gave the best performance both for description of the chromatographic shapes of individual components and for resolution of overlapping profiles. The results reported are in agreement with a previous study on unresolved peaks and represent a considerable extension in terms of analytical significance of the usefulness of the log-normal function for overlapping peak deconvolution. © 1989.