The Program Evaluation and Review Technique (PERT) model is one of the most popular methods for estimating software size and, sometimes, development effort. It mainly relies on expert judgment to estimate the ultimate size of a project. However, the basic assumption of this technique may be incorrect. The issue addressed in this paper concerns the weakness of the PERT model and proposes applying Erlang distribution to tackle this problem. An example has been included to show the effectiveness of this new model.