A Monte Carlo method for studying farm planning problems is developed. The method allows the building of a flexible model including several objective functions, economies of scale, integer formulations and interactions between activities arising from, for example, rotational yield effects. The maximum number of activities in the solutions can be determined in advance. Minimum levels for the activities can be specified to eliminate levels of no practical interest. Many solutions are obtained between which the decision-maker can choose. A statistical analysis of the solutions for different levels of the objective functions gives an informative survey of the possible plans. Decision-making based on the Monte Carlo output may be of special interest with regard to "the behavioural theory of the firm".