Nonaqueous reversed-phase high-performance liquid chromatography of 10 homogeneous triacylglycerol molecular species (TAG), both saturated and unsaturated, is carried out. The eluate from the column is detected by an ultraviolet diode array detector (DAD) on-line with an evaporative light-scattering detector (ELSD). The retention parameters (as selectivities, alpha) for 220 TAGs are determined, and the obtained values are related to the following structural parameters: total carbon number; mono-, di-, and triunsaturated fatty acid residues number/molecule; and monounsaturated fatty acid carbon number. Multiple regression analysis is carried out to obtain a relationship for the prediction of alpha values of any TAG when the same experimental conditions are used. In regard to the quantitative analysis of the separated TAG species, the dependence of response of the two on-line detectors on the aforementioned structural parameters is studied. Three different wavelengths (205, 210, and 215 nm) are considered for TAG detection by DAD; in each case, the obtained multiple regression model shows a good correlation between the dependent variable and predictive values of the TAG species (response factors and considered structural parameters, respectively). The ELSD gives responses exponentially related to injected amounts. Also, in this case, an attempt to relate the response factors of each considered detector to some structural parameters of TAG species is carried out. The results of this study are used to analyze the TAG fraction from an olive oil.