Soil contamination by heavy metals considers as main environmental problem in the world, most of them have toxic effects on plant and microorganisms in soil when permissible concentration levels are exceeded. The objective of the current work  is detection of soil contamination in southern Port-said Governorate using spectroscopy Vis-NIR (350-2500 nm). To predict contaminated soil through Vis-NIR  58 different sites were selected in southern Port-said Governorate. Spectroradiometer ASD was used to measure the spectral reflectance of soil site. Three heavy metals Pb, Mn and Cd have been selected to achieve this purpose. Stepwise multiple linear regression (SMLR) was used to construct calibration models. The concentrations of heavy metals were estimated with high accuracy where, the models were validated based on the independent validation. The results illustrated that, R² was recorded 0.94, 0.70 and 0.66 for Cd, Mn and Pb respectively. The concentration of soil contamination by Cd, Mn, and Pb are under the threshold. except area contaminated with Cd is (10791.08) faddans. However, Remote Sensing and Geographic Information System (GIS) are provide detailed spatial information on soil contamination , it's rapidly and inexpensive.