Background: The JAK1/JAK2 inhibitor RUX has demonstrated rapid and durable improvements in splenomegaly and symptoms and improved survival in the phase 3 COMFORT studies in pts with MF. Here we report final long-term safety and efficacy results after 5 years (y) of RUX treatment in COMFORT-I. Methods: In COMFORT-I, 309 pts were randomized (1:1) to RUX or placebo (PBO). RUX starting dose was based on baseline platelet count (100–200×10⁹/L: 15 mg BID; > 200×10⁹/L: 20 mg BID). Pts receiving PBO could crossover to RUX after the primary analysis (when all pts completed week [wk] 24 and half completed wk 36) or at any time if they had prespecified worsening of splenomegaly. The primary endpoint was the proportion of pts achieving ≥ 35% reduction in spleen volume (SV) at wk 24. Overall survival (OS) was estimated by Kaplan-Meier analysis according to randomized treatment. Results: Median follow-up was 268 wk at the time of this analysis. Of 154 pts randomized to PBO, 111 crossed over to RUX; median time to crossover was 41.1 wk. At wk 24, pts originally randomized to RUX had a mean SV reduction of 31.6%; this was durable for pts who continued on RUX (mean SV reduction at wk 264, 37.6%). Median duration of ≥ 35% SV reduction was 168.3 wk (range, 107.7–NE). OS favored RUX (HR 0.69; 95% CI: 0.50, 0.96; P= 0.025), with 69 and 82 deaths among pts originally randomized to RUX and PBO, respectively. Median OS has not been reached for RUX. Mean platelet count and Hgb initially decreased through 3 mo. Mean Hgb gradually increased toward baseline. After wk 24, mean Hgb and platelet count generally remain stable through 5 y. New onset grade 3 or 4 anemia was 25.2% for the RUX arm and 26.1% for PBO crossover; grade 3 or 4 thrombocytopenia occurred in 12.3% and 13.5% of pts, respectively. Notable AEs included herpes zoster (10.3% and 13.5% of RUX and PBO crossover pts, respectively); basal cell carcinoma (7.7% and 9.0%, respectively); acute myeloid leukemia (5 pts in each arm). Conclusions: After a median follow-up of 268 wk, the hazard ratio for OS favored pts randomized to RUX over those randomized to PBO, and SV reductions were sustained with long-term therapy. Collectively, these data support the durable efficacy and long-term safety of RUX in pts with MF. Clinical trial information: NCT00952289