OBJECTIVES: To determine the cost-effectiveness of prophylactic administration of Filgrastim compared with no use, during the induction phase of chemotherapy in adults with Acute Lymphoblastic Leukemia (ALL) in the Colombian context. METHODS: A decision tree with a time horizon of 30 days is built under the third-party payer perspective including only direct costs. The costs of procedures and medications were taken from official sources and an institution of national reference of oncology services. The safety and effectiveness data were taken from the literature and two Colombian cohorts (one retrospective and one prospective) with patients older than 15 years. The unit of outcome was the proportion of deaths averted. The incremental cost effectiveness ratio (ICER) was estimated, univariate sensitivity and probabilistic analysis were performed. RESULTS: Model results indicate that under the scenario of a  clinical trial not using factor was a dominated alternative (ICER of - 61,753,681 COP per death averted). In contrast, using data from the Colombian cohorts, factor was dominated strategy (ICER of - 141,421,622 COP for retrospective cohort and prospective cohort -215,449,438 COP). The variable that most impacted the outcome was the incidence of febrile neutropenia (12% for the clinical trial, 60%  retrospective cohort and 83%  prospective cohort). The results were robust to the probabilistic sensitivity analysis. With the data from the clinical trial in 94% of cases using factor was cost effective, while in the Colombian data in 84% and 72% of cases (retrospective and prospective cohort respectively) was not cost effective to use factor. CONCLUSIONS: With Colombian information the prophylactic use of the factor under chemotherapeutic induction in adults with ALL turns out to be not cost-effective. The gap in the results suggests a careful extrapolation of information from clinical trials (ideal world) to develop economic evaluations in Colombia, and its impact on decision making.