Octane number is an important characteristic of spark engine fuels such as gasoline and jet fuel or fractions that are used to produce these fuels (i.e., naphthas) and its represents antiknock characteristic of a fuel. It is that percentage of isooctane in a blend of isooctane and normal heptane that exactly matches the knock behavior of the gasoline. Two octane numbers are usually determined. The first is the research octane number RON and the second is the motor octane number MON. The main objective of this study is to predict the octane numbers (RON & MON) using various correlations methods for the petroleum fractions of light and heavy naphtha. The study was conducted on the petroleum fractions of both light and heavy naphtha for the blend of Messla and Sarir crude oils at Tobruk refinery. The main objective of the study is to highlighting on procedures for prediction of ON using approaches correlations. The results show that the estimation of RON by using RON values for pure hydrocarbons shows 73.93 with a deviation 2.43 from clear RON 71.50; while, estimation RON from the pseudocomponent method using predicted PNA composition is 74.12 with an error 2.62 from clear RON 71.50. RON from Nelson graphical correlations using K_w , T_b and total paraffins (x_p, x_Ip) and T_b reported 73.98 and 73.90 with an errors 2.48 and 2.40 from clear RON 71.50 respectively. The calculated MON from actual reported RON shows 〖(MON)〗_(est.) = 67.55 with an error 1.43 from the reported value 68.98; and from predicted RON shows 〖(MON)〗_(est.) = 69.56 with an error -0.58, is in good agreement with the reported value 68.98. Consequently, the findings indicate that the approaches correlations can be applied to predict the octane numbers for petroleum fractions because of their consistency with the reported experimental data.