This chapter aims to generate non-stationary arrivals in Arena. The Arena Input Analyzer allows both discrete and continuous empirical distributions to be formed using the Fit/Empirical option. The Arena software incorporates an Input Analyzer software tool, which provides facilities to derive theoretical and empirical distributions. The ability to generate non-stationary arrivals is important because many manufacturing and service systems experience arrival rates that vary dramatically over time. A probabilistic model for this, the non-stationary Poisson process, can be used to reflect time-varying arrival patterns in an accurate manner. To deal with time-varying patterns, we can identify periods over which the arrival rate is fairly flat and then count the number of arrivals in each of these slots. The schedule will repeat when the last duration has occurred unless the last duration is specified as Infinite. The use of the create and schedule modules to create non-stationary arrivals is also shown.