OBJECTIVES To estimate the total costs of arsenic trioxide and all-trans retinoic acid (ATO+ATRA) versus ATRA and idarubicin (AIDA) regimens in Italy when used in 1^st-line APL treatment.  ATO+ATRA is approved in 2^st-line treatment, but commonly used globally in 1^st-line. METHODS A Markov model was developed with three health states: non-progressive disease, progressive disease and death.  Each month, patients could move from non-progressive to progressive disease or die from either state.  After progression, patients discontinued treatment and switched to the other regimen.  Treatment regimens, efficacy and adverse events were derived from published sources and expert opinion, while each arm’s unit costs (induction and consolidation for both; maintenance for AIDA only) were collected from standard Italian sources.  Per-patient costs were reported, and extensive one-way sensitivity analyses were conducted. RESULTS Expected 2-year pharmacy costs for ATO+ATRA were €46,600 versus €5,300 for AIDA.  However, direct medical (DM) costs (e.g., monitoring, hospitalizations, etc.) for ATO+ATRA were €11,300 versus €28,500 for AIDA.  The higher costs stemmed from AIDA consolidation taking place in a hospital setting and 2-year maintenance monitoring costs. The treatment failure likelihood for patients on AIDA was 14% versus 3% on ATO+ATRA.  As these patients switched to 2^nd-line treatment, relapse costs for AIDA were €1,500 compared to €400 for ATO+ATRA. AIDA patients incurred higher costs from adverse events (AEs) than ATO+ATRA (€600 vs. €300, respectively). Results were most sensitive to consolidation cost changes:  ATO+ATRA pharmacy costs and AIDA DM costs. CONCLUSIONS The results suggest potential cost savings due to DM, progression and AE costs for ATO+ATRA, although AIDA pharmacy costs are lower.  Additionally, ATO+ATRA patients have lower clinical AE risks and shorter treatment duration (8 months versus 28 months for AIDA), which could lead to patients’ improved quality of life and cost savings.  Further research is needed to determine the cost-effectiveness of ATO+ATRA.