OBJECTIVE: To conduct a retrospective drug utilization evaluation comparing the use of granulocyte colony-stimulating factor (G-CSF) at a university medical center with the American Society of Clinical Oncology (ASCO) CSFs Practice Guidelines. METHODS: Patients who received G-CSF from June 1, 1996, to December 31, 1996, were identified through the pharmacy computer system and the medical records were reviewed for a randomly selected sample of 26% of the 289 patients identified. Outpatient, inpatient, and electronic medical records were reviewed for the indication, dosage, day of initiation, day of discontinuation, and absolute neutrophil count (ANC) monitoring plan for each course of G-CSF; these records were subsequently compared with the ASCO guidelines. RESULTS: The use of G-CSF after chemotherapy was evaluated in 51 patients who received a total of 182 courses of G-CSF. The goal of chemotherapy was curative in 61% of courses. Sixty-five percent of G-CSF courses were prescribed for primary prophylaxis. Of these, 74% followed chemotherapy in patients with an expected incidence of febrile neutropenia ≥40% or followed chemotherapy in patients with compromised marrow reserve secondary to extensive prior therapy or in patients older than 60 years. Most of the G-CSF courses (75%) were rounded to the nearest vial size. The areas of greatest departure from the ASCO guidelines included aspects of initiation and discontinuation of G-CSF courses and inadequate documentation of ANC recovery. CONCLUSIONS: These results demonstrate a number of specific opportunites for oncology pharmacists to improve the use of G-CSF in patients receiving chemotherapy. Recommendations were made to the pharmacy and therapeutics committee and medical oncologists to improve compliance with the ASCO guidelines.