Crop modelling has had a mixed press over the years, driven as much by the
community's expectations as by the targets identified by modellers. This sh
ort review gives a personal perspective on the scope and value of models of
crop processes and performance, in the context of the increasing demand to
make predictions about the best course of action for farmers. The goal for
these predictions can range from the best rate or timing for a pesticide s
pray to the likely pressure on the farming community if predicted levels of
climate change are realised. The primary conclusion is that a clear statem
ent of objectives is the most important precursor to the use of models of c
rop processes or systems.