Crude oil price forecasts are an important component of sustainable development of many countries as crude oil is an unavoidable product that exist on earth. In this paper, a model based on markov model and artificial neural network for crude oil price forecasting was developed and their relative performances were compared using SEM (AMOS). Four different error analysis techniques were employed to evaluate the most accurate model. Path analysis of structural equation modelling was used to buttress the findings of the error analysis, path analysis modeled the relationships of the forecasted prices and the actual crude oil price in order to get the most accurate forecast. The key variables used to develop the models were monthly crude oil prices from PETRONAS Malaysia. The markov model were found to provide more accurate crude oil price forecast than the artificial neural network. The results of this study indicate that markov models are a potentially promising method of crude oil price forecasting that merit further study.