PART is a highly automated planning system in which both process and production planning functions are integrated. This paper discusses a method to improve machine tool selection in process planning by integration with loading. A method is presented to select the best process plan from a number of possible alternatives taking into account the limited availability of resources. Various aspects of the quality of a process plan are evaluated and expressed in the so-called evaluation time. To prevent redundant work, partly worked out process plans are considered as alternatives. The consequences of the different alternatives have to be estimated which includes the estimation of machining times. The loading problem is modelled as the minimization of the total evaluation time for a given order set, subjected to capacity constraints.