Over the past two decades deaths from opioids and other drugs have grown to be a major U.S. population health problem, but the magnitude of the crisis varies across the U.S., and explanations for widespread geographic variation in the severity of the drug crisis are limited. An emerging debate is whether geographic differences in drug mortality rates are driven mostly by opioid supply factors or socioeconomic distress. To explore this topic, I examined relationships between county-level non-Hispanic white drug mortality rates for 2000-02 and 2014-16 and several socioeconomic and opioid supply measures across the urban-rural continuum and within different rural labor markets. Net of county * demographic composition, average non-Hispanic white drug mortality rates are highest and increased the most in large metro counties. In 2014-16, the most rural counties had an average of 6.2 fewer deaths per 100,000 population than large metro counties. Economic distress, family distress, persistent population loss, and opioid supply factors (exposure to prescription opioids and fentanyl) are all associated with significantly higher drug mortality rates. However, the magnitude of associations varies across the urbanrural continuum and across different types of rural labor markets. In rural counties, economic distress appears to be a stronger predictor than opioid supply measures of drug mortality rates, but in urban counties, opioid supply factors are more strongly associated with drug mortality rates than is economic distress. Ultimately, the highest drug mortality rates are disproportionately concentrated in economically distressed mining and service sector dependent counties with high exposure to prescription opioids and fentanyl. JEL Codes: I1, I3, J21, K32, R1.