Among the poor in developing countries, up to 20% of an infant's life experience may include diarrhea. This problem is spatially related to the lack of potable water at different sites. This project used risk analysis (RA) methods and geographic information system (GIS) technologies to evaluate the health impact of water source. Maps of Imo State, Nigeria were converted into digital form using ARC/INFO GIS software, and the resulting coverages included geology, hydrology, towns, and villages. A total of 11,537 diarrheal cases were reported. Thirty-nine water sources were evaluated. A computer modeling approach called probabilistic layer analysis (PLA) spatially displayed the water source at layers of geology, hydrology, population, environmental pollution, and electricity according to a color-coded five-point ranking. The water sources were categorized into A, B, and C based on the cumulative scores < 10 for A, 10-19 for B, and > 19 for C. T-test showed revealed significant differences in diarrheal disease incidence between categories A, B, and C with mean +/- SEM values of 1.612 +/- 0.325, 6.257 +/- 0.408, and 15.608 +/- 2.151, respectively. The differences were significant between categories A and B (P = 0.0000022), A and C (P = 0.0000188), and B and C (P = 0.0011348). The PLA enabled estimation of the probability of the risk of diarrheal diseases occurring at each layer and solutions to eliminate these risks.