Abstract
A new and simple quantitative structure-retention relationship (QSRR) was tested as a predictive model for adjusted retention times in complex petroleum condensate fractions. This relationship adopted the form of a non-linear collective retention-variables model. The adjusted retention times were correlated with the components molecular descriptors, e.g. total path counts and boiling temperatures, by multi-linear regression analysis. The obtained two QSRR models show an acceptable predictive accuracy with R
2 of 0.9949 and 0.9856, respectively. Stability and validity of the models were tested by comparing the calculated and the experimental retention indices.