Hospital costs relate significantly to LOS which is often used as a surrogate for costs in measuring the success of some cost containment programs. This study was designed to determine if yearly decreases in LOS for Pediatric AuBMTS in one hospital have been due to improved therapies, shorter preparative regimens, or better discharge planning. METHODS: Retrospective data including LOS were collected on 29 AuBMT patients discharged alive in FY's 1993-1996. Hospital days were stratified by fiscal year & divided into 3 phases; 1) From admission to date of BMT; as a measure of of the preparative therapy phase (PREBMT), 2) From date of BMT to the day the neutrophil count recovered to 500/cmm; as a measure of therapeutic efficacy (BMT>500) and 3) From BMT>500 to Discharge; as a measure of discharge planning (500-DC). For bivariables, data were grouped into 1993-94 and 1995-96 subsets and tested using x2 and t-tests. Year specific data were used in linear regressions for multivariate analyses. RESULTS: Mean (±SE) 39 ±3 day LOS in 1995-1996 was less than 53 ±3 stay in 93-94 (p=.004). By phase, PREBMT fell from 10 ±.6 to 8±.3 days (p<.03), BMT>500 from 29 ±3 to 20 ±2 days(p<.04), and for 500-DC insignificantly from 13 ±2 to 11 ±2 days (p=ns). In multivariate analysis, with LOS as the dependent variable, BMT>500 accounted for 73% of variance, 500-DC 23%, PREBMT 3%, and Fiscal year <1%. CONCLUSIONS: Apparently in this hospital, LOS for BMTS decreased because of better therapies shortening the time to an ANC of 500 and only secondarily because of changes in length of the preparative regimen. Although contributing significantly to LOS variance, no significant contribution as yet was attributable to discharge planning. Advances in medical therapeutics influence LOS and need incorporation when analyzing cost containment programs.