Gasoline fuels are complex mixtures which consist of more than 200 different hydrocarbon species. In order to decrease the chemical and physical complexity, oxygenated surrogate components were used to enhance the fundamental understanding of partially premixed combustion (PPC). The ignition quality of a fuel is measured by octane number. There are two methods to measure the octane number: research octane number (RON) and motor octane number (MON). In this paper, RON and MON were measured for a matrix of n-heptane, isooctane, toluene, and ethanol (TERF) blends spanning a wide range of octane number between 60.6 and 97. First, regression models were created to derive RON and MON for TERF blends. The models were validated using the standard octane test for 17 TERF blends. Second, three different TERF blends with an ignition delay (ID) of 8 degrees for a specific operating condition were determined using a regression model. This was done to examine the model accuracy for ID and study fuel composition effect on combustion events and emissions.The results showed a good agreement between predicted and tested RON and MON with an accuracy of ±0.6. The model also had high accuracy during extrapolation for some fuel blends. For toluene and reference fuel blend (TRF) the model was more accurate for MON than for RON, while the situation was the opposite for ethanol and reference fuel blend (ERF) i.e. the model worked better for RON than for MON during extrapolation. The ignition delay was similar for all three TERF blends despite the differences in their composition. However, high concentration of toluene resulted in higher levels of HC, NOx, and smoke emissions.