This article presents original research findings to illustrate the influence of surveillance definitions on infection rate. Several definitions for urinary tract infection were applied to clinical and laboratory data from rehabilitation hospital patients, and the effect on sensitivity and specificity examined. We demonstrate that it is essential to take into account the influence of the surveillance definition when interpreting surveillance data, and the importance of ensuring that the surveillance definition is compatible with the surveillance objectives when selecting a definition. Surveillance definitions that serve well in one surveillance program may not serve well in another.