We propose an intelligent approach to gasoline price forecasting as an alternative to the statistical and econometric approaches typically applied in the literature. The linear nature of the statistics and Econometrics models assume normal distribution for input data which makes it unsuitable for forecasting nonlinear, and volatile gasoline price. Karhunen-Loève Transform and Network for Vector Quantization (KLNVQ) is proposed to build a model for the forecasting of gasoline prices. Experimental findings indicated that the proposed KLNVQ outperforms Autoregressive Integrated Moving Average, multiple linear regression, and vector autoregression model. The KLNVQ model constitutes an alternative to the forecasting of gasoline prices and the method has added to methods propose in the literature. Accurate forecasting of gasoline price has implication for the formulation of policies that can help deviate from the hardship of gasoline shortage.