Petroleum contributes significantly to economic growth, but pollution near oilfields poses risks to ecosystems and human health. The soil petroleum content and sources in the Yanchang oilfield, located in northern Shaanxi, remain unclear and present an environmental threat. To address this, a comprehensive soil contamination survey was conducted in the surrounding area. The total petroleum hydrocarbons (TPH) content in the soil exhibited notable spatial variability, ranging from 102.06 to 4290.03 mg/kg, with short-chain components (C10-C40) predominating and concentrations inversely correlated with the distance from the petroleum wells. A multilayer perceptron prediction model identified oil volatilization and migration as key factors influencing oil concentration. Additionally, sixteen polycyclic aromatic hydrocarbons (PAHs) were detected, with total concentrations (Sigma 16 PAHs) ranging from 216.48 to 3637.59 mu g/kg, and chrysene being the most abundant. Most soil samples exhibited PAH levels exceeding the toxic equivalent concentration, indicating significant contamination and biotoxicity risks. Source analysis revealed that oil spills and incomplete biomass combustion were the primary contributors to PAH contamination. These findings provide a crucial scientific foundation for soil pollution management and public health protection.