A dynamic probabilistic model has been designed to determine the technical and economic consequences of various biological variables and management strategies concerning reproduction, replacement and calving patterns in dairy herds. The Markov chain approach is used to simulate herd dynamics. Herds are described in terms of states animals can be in and the possible transitions between the states. The corresponding transition probabilities are derived from input values concerning biological variables and management strategies. The model has the property of calculating easily the herd structure at equilibrium (steady state). Sets of input values can be evaluated by comparing the technical and economic results of the corresponding herds at equilibrium. Moreover, herd dynamics can be studied over a period of time to gain insight into how equilibrium is reached. Input values can be modified to be farm-specific. The model is used to compare different calving patterns of the herd under different circumstances. Results for Dutch conditions show that it is most profitable to have calvings concentrated in autumn.