This paper applies non linear methods to analyze and predict the weekly Gas and Oil prices. The aim of the analysis is to quantitatively show if the corresponding time series are deterministic chaotic and if they are predictable. The research employs Grassberger and Procaccia methodology into the time series analysis in order to estimate the correlation and minimum embedding dimensions of the corresponding strange attractor. In order to achieve out of sample one week ahead prediction, a local weighted least squares over all neighbors’ projections k-steps into the future has been applied. More specifically, the gas time series presents persistence behavior with max Lypapunov exponent to be 0.378, indicating that a prediction horizon is about two steps ahead. Almost similar results has been obtained with the time series of crude oil .