In this paper we treat a general worst-case system identification prob
lem. This problem is worst-case with respect to both noise and system
modeling uncertainty. We consider this problem under various a priori
information structures. We determine bounds on the minimum duration id
entification experiment that must be run to identify the plant to with
in a specified guaranteed worst-case error bound. Our results are algo
rithm independent. We show that this minimum duration is prohibitively
long. Based on our results, we suggest that worst-case (with respect
to noise) system identification requires unrealistic amounts of experi
mental data.