The Poisson process has been widely used in the literature to model call center arrivals. In recent years, however, there have been empirical studies suggesting the call arrival process has significant non-Poisson characteristics. In this paper, we introduce a new doubly stochastic Poisson model for call center arrivals and develop a Bayesian approach for the parameter estimation via the Markov chain Monte Carlo method. The model can well capture the call arrival process as illustrated by a case study.