Among resources considered in the farm system labour and machinery dominates all other cost categories and much is to be gained by adapting and operating these factors within the boundaries of the actual needs arising from farm size, crop plans etc. A model supporting the process of choosing the optimal machinery sizes is developed, and is described in this paper. The optimization model is a non-linear programming model implemented by using the programming software GAMS. It is based upon a least-cost concept involving all fixed and variable costs (including timeliness costs) for a particular farm size and crop plan. The output provided by the model is the size of each machine, needed tractor power and number of tractors. Also, the effective capacities related to machinery sets and the dispersion in time of the performance of each operation are put forward, The selection is based upon a farm-oriented matrix involving various types of constraints, such as available man, machine and tractor hours, timeliness and workability of operations, agronomic window and sequence of operations, etc. The model is to be used as a decision support tool both when analysing different farm machinery systems separately, and as an integrated part of an overall farm planning model currently under development by the effort of a multi-disciplinary research project. In the last-mentioned case the model is used when specifying the initial conditions and expectations prior to a dynamic simulation of the farm development over a number of years.