Most applications of Wireless Sensor Networks (WSNs) the sensors are deployed on unfavorable environment such as chemical reactor or battlefield that with high temperature, noise, and interference, could probably incur sensor nodes sense, compute, or communicate improperly. Those also raise error responds to the data collectors. In this paper, a fault estimation model is proposed and can be used to construct the fault map which can be used in WSNs widely, especially in hostile environments. Sensor nodes are transmitting environmental data when it detected with some extra sensed data, such as outward temperature and noise. The fault probability can be estimated and recognized with this extra information using Bayesian Belief Network (BBN). In this paper, we develop a fault-estimation algorithm is based on the cluster-based framework and show how this model works to find out the probably faulty nodes. Then we simulate and construct the fault map of WSNs. Keywords: fault estimation, fault map, Bayesian Belief Network, wireless Sensor network