The frying-lives of olive oil (OO), sunflower oil (SO), and a blend of these oils (BO) were assessed and compared by measuring the polar content (PC) and the TAG oligomer content (TOC) in the oils. Oil was replenished with fresh oil every 10 uses in all three oils. Changes in the PC and TOC in relation to the number of frying uses were fitted to different curvilinear models. The power model yields the highest R 2 value in the three oils. The 25% PC was surpassed after 32.2 fryings [95% confidence interval (95% CI) 29.1-36.8] in the OO, 22.5 (95% CI, 21.0-24.6) fryings in the SO, and 27.5 (95% CI, 25.5-30.1) fryings for BO. However, according to the 10% TOC cutoff point, OO should be discarded after 25.0 (95% CI, 22.8-28.7) fryings, SO after 15.0 (95% CI, 14.8-15.4) fryings, and BO after 17.7 (95% CI, 17.0-19.1) fryings. Nevertheless, changes in PC and TOC were different only between OO and SO (P < 0.05 and P < 0.02, respectively), indicating that OO performs better than SO, and that BO can be used as an alternative when both frying-life and price are under consideration. Present findings suggest the need to unify criteria in different countries for oil disposal because the 25% PC corresponds to a TOC higher than 10%.