Our study is the first to provide statistics on opioid use among U.S. Social Security Disability Insurance (SSDI) applicants. We use an innovative machine-learning method to identify opioids in open-ended text fields in SSDI administrative data. We find that more than 30% of applicants between 2007 and 2017 reported using one or more opioids, a rate that is about 50% higher than in the general population. Rates of reported opioid use varied over time, peaking at 32% in 2012. Reported opioid use also varied by age, gender, education, receipt of SSI, and across states. There was a positive and statistically significant association between (a) reported opioid use and SSDI awards and (b) reported opioid use SSDI award and death; these are associations and do not demonstrate a causal relationship.