Flood hazards have been overwhelming in recent years. The hazard tends to impact on human lives and resulting to huge economic damage across the World. However, the prediction of the magnitude of flood has been faced with challenges such as inaccurate data, poor assessment of drainage basin, river pollution and encroachment, especially in Nigeria across the river areas. In this study, Geographical Information System (GIS) was used to derive cross-sectional variables in order to complement the prediction of magnitude of flood along Foma river areas. The Foma river was buffered 15 and 30 meters to expose 49 and 105 structures which were highly and fairly vulnerable to flood hazards respectively out of 530 structures captured. However, the remaining 376 structures were classified not vulnerable to flood hazard along Foma river areas. The accuracy of the ordered logit model was 81%, while the classification error due to the harmonization of the precision value (0.8026) and the recall value (0.6386) was put at 10%. In addition, the cross-sectional variables that were found to be significant at α = 0.005% are the river watersheds, the vulnerability status classification of structure across the river areas, the vulnerable structures identified, inadequate bridges and culverts along the river areas, inappropriate size of bridges and culverts, and extreme pollution along the river areas. This study is recommending the use of significant cross-sectional variables to complement the prediction of the magnitude of flood along the river areas.