Abstract-In this paper, we proposed the Markov Chain Approach (MCA) to evaluate the average run length (ARL) of exponentially weighted moving average (EWMA) control chart when zero-inflated counted are observed in a negative binomial model. Furthermore, the efficiency of the MCA is compared with Monte Carlo Simulation via the CPU times which the former is much saving computational times used when compared with the latter method.Index Terms-Exponentially weighted moving average, zero-inflated negative binomial, average run length, markov chain, computational times.