Models are used in many policy arenas to predict the future consequenc
es of current decisions. A model is typically viewed as a rational, ob
jective means of processing complex information to predict future cond
itions. With respect to socioeconomic modeling, policy scientists have
found that models frequently do not live up to these promises - they
often incorporate the modelers' biases, are difficult for decision mak
ers to use, and do a poor job of point predictions. While awareness of
these characteristics of socioeconomic models has increased, less att
ention has been paid to such factors in environmental modeling. This p
aper explores the implications of policy scientists' observations abou
t socioeconomic models for the use of water quality models in environm
ental decision making. For example, the relatively simple task of mode
ling for an estuarine waste load allocation decision incorporates judg
ment in model choice, calibration, and use. More complex watershed mod
els involve even more choices that have implications for decision maki
ng. Thus, environmental models are not strictly technical inputs to th
e policy process. Model users, regulators, and the public should be aw
are of the judgments and uncertainties involved in surface water quali
ty modeling so that model results are used appropriately in the decisi
on-making process.