PURPOSE: Patients with advanced cancers frequently experience pain. Opioids are commonly prescribed to treat cancer-related pain, but their use might be associated with undesirable consequences including adverse effects and tumor progression, resulting in increased heath care utilization and shorter survival. We examined these possibilities in a large cohort of patients diagnosed with ten common advanced malignancies. METHODS: We identified 1386 newly diagnosed patients with stage IV non-hematologic malignancies from 2005 to 2013 and ascertained opioid utilization within 90 days of starting anti-cancer treatment using electronic medical record and tumor registry data. Opioid utilization was stratified into low opioid (LO; < 5 mg oral morphine equivalents (OME)/day) and high opioid (HO; >/= 5 mg OME/day). Health care utilization included tallies of emergency room, urgent care, and inpatient visits. The association of opioid use, tumor type prognosis, age, and gender with overall survival was analyzed in univariate and multivariate models. RESULTS: HO use patients (n = 624) had greater health care utilization compared to LO use patients (n = 762; p < 0.05). HO use patients also had shorter survival (median survival, 5.5 vs 12.4 months; p < 0.0001). On multivariate analysis, HO use remained associated with shorter overall survival (HR 1.4; 95% CI, 1.3-1.6; p < 0.0001) after adjusting for age, gender, and prognostic group. CONCLUSIONS: In advanced cancer patients, HO use is associated with greater health care utilization and shorter survival. Prospective studies using opioid-sparing approaches are indicated, to confirm these retrospective findings and to evaluate if these undesirable effects associated with opioid use can be mitigated.