Recent involvement in designing and developing a simulation model to allow interaction between biophysical, economic and social processes has also led to interest in uncertainty and error propagation in models. This uncertainty exists in each of the biophysical, economic and social domains. With regard to the hydrologic processes there appears to be an indeterminacy principle that makes up-scaling difficult. When this is combined with the uncertainty in the other aspects of the model it would suggest that caution is needed in interpreting output from such models. The uncertainty in biophysical, social and economic systems is a combination of uncertainty in; data input and model structure. As we build models where non-linearities due to the model structure are incorporated the relative uncertainty in the outputs will grow rapidly. Using a simple model where data input errors are either added or multiplied together we can see the consequences for the relative error in the output (figure 1). Suppose we have a process that results in a local (cell) parameter yi with standard deviation σi and the area of the plot is ai. We will assume that the same value and standard deviation occur in n other plots such that: The uncertainty present in models requires that when using the results of models that clients are made aware of the extent of these errors and their nature. We feel that policy makers should be made aware of the uncertainty when using such models. Decisions still need to be made and modelled scenarios provide inputs that help in make sense of the studied system. However, having made a decision to change a system monitoring of the results is required to determine if the desired response has occurred. Indeed, the dynamic between the uncertainties contained with the model and those in the minds of its clients is itself a social process that can be monitored and managed.