Nearly 100,000 deaths every year in the United States are now wholly or in part attributed to opioid overdose. Innumerable people have died from the indirect effects of opioids, such as heart disease exacerbated by long-term opioid misuse, and the degradation of protective social supports. Dual-eligible Medicare and Medicaid beneficiaries under the age of 65 (nonelderly duals) comprise a population with serious risk factors for premature death following a nonfatal overdose. By definition, they have disabling conditions but cannot afford their health care. Nonelderly duals also may have beneficial relationships with health care providers as a result of managing their comorbidities. These relationships may provide opportunities to initiate medication for opioid use disorder outside emergency department encounters. The goal of this dissertation is to describe the relationship between nonfatal opioid overdose, health care utilization, and 12-month mortality in this population. I used Medicare and Medicaid claims data to examine these associations among nonelderly duals who survived an opioid overdose between 2014-2016. Data from 2013-2017 were used to capture health care utilization and diagnoses prior to and following the index overdose.Paper 1 describes the epidemiology of nonfatal opioid overdose and 12-month mortality among nonelderly duals who overdosed in the study period. Nearly 1 in 9 nonelderly duals who experience nonfatal opioid overdose died in the following year. Sex and some observed comorbidities were predictive of post-overdose mortality. Medication for opioid use disorder (MOUD) was associated with reduced mortality, but was rarely indicated in this population. Among those who died, the average time to death was 5 months, and most beneficiaries visited health care providers prior to death, suggesting there may be opportunities following the overdose to prevent death. Paper 2 reports associations between MOUD and 12-month mortality among nonelderly duals as assessed using propensity score methods. Associations were described for the overall population as well as by sex and general level of health. Active MOUD at the time of the nonfatal overdose was associated with lower rates of 12-month mortality among men and healthier beneficiaries only. These findings underscore the importance of addressing differences in subpopulations that may affect access to and effectiveness of MOUD. Paper 3 describes health care utilization and its association with 12-month mortality by sex and diagnosis of schizophrenia. Men without schizophrenia had higher mortality rates than other groups in this study, despite having the highest rate of indicated MOUD use. Additionally, women had more indicators of serious chronic illnesses. Emergency department visits were not associated with death within 12 months, and inpatient visits were only associated with mortality among beneficiaries with schizophrenia. Certain physical conditions, such as congestive heart failure, were associated with 12-month mortality only among beneficiaries without schizophrenia, suggesting that physical comorbidities have differential effects between groups. On average, beneficiaries in all groups saw outpatient health care providers on 24 or more days in the year prior to the OD. These visits to health care providers in familiar settings may provide opportunities to initiate MOUD other than at the index overdose event.