This paper uses Monte Carlo Simulation to evaluate the seismic hazard in two nuclear power plants sites in Taiwan. This approach is different from the two commonly-used methods, featuring the direct use of observed earthquakes to develop magnitude and distance probability functions. Those earthquakes with larger sizes and closer to the site are considered capable of causing damage on engineered structures. The result shows that even though two NPP sites in Taiwan are only 30 km away from each other, NPP site 4 is situated with the seismic hazard four-time as large as NPP site 1. Given our limited understanding and the complicated, random earthquake process, none of a seismic hazard analysis is perfect without challenge. The accountability of seismic hazard analysis relates to analytical transparency, traceability, etc., not to analytical complexity and popularity.