OBJECTIVES: Our recent study showed that multiple factors including patient characteristics, comorbidities, prior medications, prior health care utilization and costs were associated with initial prescription dosage of duloxetine in the treatment of major depressive disorder (MDD). The purpose of the present study was to examine the associations between initial prescription dosage of duloxetine and hospitalization and costs in the 1-year follow-up period. METHODS: Of 10,128 patients with MDD in a large commercial administrative claims database who were initiated on duloxetine between July 1, 2005 and June 30, 2006 and had no prescriptions of duloxetine in the prior 6 months, 6132 (60.5%) had continuous enrollment for 12 months prior to and post-duloxetine initiation. According to initial prescription dose, patients were divided into 3 cohorts: low-dose (<60 mg/day, n=1,989), mid-dose (60 mg/day, n=3,733), and high-dose (>60mg/day, n=410). Intent-to-treat analysis was conducted to compare health care utilization and costs in the 1 year prior and post-initiation across the 3 cohorts. RESULTS: During the post 1 year, hospitalization rate was significantly reduced by 5.7% for mid-dose cohort (p<.001), 6.9% for high-dose cohort (p <0.05), and 7.3% for low-dose cohort (p<0.001). Pharmacy costs in the post 1 year increased in all 3 cohorts: low-dose, $672 (p<.001); mid-dose, $825 (p<0.001); and high-dose, $1437 (p>0.05). However, mean medical costs declined in high-dose patients ($2617) and mid-dose patients ($340), but increased in low-dose patients ($261). Mean total health care costs numerically declined by $1,180 ($20,225-$21,406) in high-dose patients, but increased in low-dose patients by $933 ($18,002–$17,069) and mid-dose patients by $485 ($16,917-$16,432), although these cost changes were not significant.   CONCLUSIONS: The hospitalization rate was significantly reduced in all patients after duloxetine initiation. Although pharmacy costs increased more in high-dose patients than in mid- and low-dose patients; the higher medication costs were offset by lower medical costs.