The purpose of this paper is to discuss the validation procedures used for validating a computer model developed for seismic hazard analysis. An elaborate and detailed computer model (J. M. Ferritto, seismic hazard analysis, users guide #UG-0027, NCEL, Port Hueneme (1993)) was developed at the Naval Civil Engineering Laboratory (NCEL), Port Hueneme, California, to perform seismic hazard analysis. The purpose of this model is to predict the probabilities of non-exceedance of a given level of acceleration through the use of cumulative density function (CDF) of the exponential distribution. This CDF is generated by determining the contributions of various faults to the overall site acceleration level through the use of recurrence relationships. The model is presently being used to predict the non-exceedance acceleration probabilities for regions with single and multiple faults. The paper deals with validation of existing mathematical models used for generating the cumulative density function (CDF). Specifically, the tool used for random analysis, namely Monte Carlo simulation, is checked and found to be consistent with the general simulation procedures used in practice. Improved methods, such as variance reduction techniques (VRT) in conjunction with traditional Monte Carlo simulation procedures for generating the CDF, are used in this study to validate/improve the existing model. In addition, the actual CDF is also generated based on the actual data (without assumption of any particular distribution) both for direct simulation as well as for VRT. All these are compared with the theoretical model used by NCEL. In all the cases, the non-exceedance acceleration probabilities predicted by the theoretical model are lower than the corresponding actual probabilities. This is checked both for single fault as well as multiple faults and the same behavior is repeated. Hence it is concluded that the NCEL model is conservative. Specific suggestions to improve the model are given in the detailed report.