Few studies have attempted to implement the 1993 A.S.P.E.N. Guidelines into clinical practice. We retrospectively studied parenteral nutrition (PN) use over a 3-month period in 148 patients. Forty-three percent of PN starts lasted < or = 5 days. This short-term use of PN was more likely to occur when peripheral access for PN was used. Overall, short-term use of PN was found over a wide variety of diagnoses for a large number of reasons given in the medical record. Next, an algorithm was developed for the decision to initiate PN based on the 1993 A.S.P.E.N. Guidelines and was used to evaluate the use of PN. Several potential means for reducing overuse of PN were identified: using established criteria for malnutrition, restricting PN to known conditions in which the gastrointestinal (GI) tract cannot be used, and recognizing the transitory nature of many GI disorders.