To detect adulteration in gasoline, an automatic distillation apparatus was set up to measure the recovered volume and temperature simultaneously. The level metering was performed by online image processing instead of the conventional visual operator-based measurement. To investigate the effect of additives in super gasoline, regular gasoline and diesel were added and the distillation curves were analyzed. The principal component analysis model was employed to reduce the obtained data. Finally, an artificial neural network was applied to predict the volume percentage of adulterants in super gasoline. Statistical analysis showed that the proposed model has a mean relative error and correlation coefficient of 4.6 % and 0.995, respectively.