The abuse of prescription drugs has increased dramatically since 1990. Persons who overdose on such drugs frequently consume large doses and visit multiple providers. The risk of fatal overdose for different patterns of use of opioid analgesics and sedative/hypnotics has not been fully quantified. Matched case-control study. Cases were 300 persons who died of unintentional drug overdoses in New Mexico during 2006-2008, and controls were 5,993 patients identified through the state prescription monitoring program with matching 6-month exposure periods. Death from drug overdose or death from opioid overdose. Exposures were demographic variables and characteristics of prescription history. Crude and adjusted odds ratios (AOR) were calculated. Increased risk was associated with male sex (AOR 2.4, 95% confidence interval [CI] 1.8-3.1), one or more sedative/hypnotic prescriptions (AOR 3.0, CI 2.2-4.2), greater age (AOR 1.3, CI 1.2-1.4 for each 10-year increment), number of prescriptions (AOR 1.1, CI 1.1-1.1 for each additional prescription), and a prescription for buprenorphine (AOR 9.5, CI 3.0-30.0), fentanyl (AOR 3.5, CI 1.7-7.0), hydromorphone (AOR 3.3, CI 1.4-7.5), methadone (AOR 4.9, CI 2.5-9.6), or oxycodone (AOR 1.9, CI 1.4-2.6). Patients receiving a daily average of >40 morphine milligram equivalents had an OR of 12.2 (CI 9.2-16.0). Patients being prescribed opioid analgesics frequently or at high dosage face a substantial overdose risk. Prescription monitoring programs might be the best way for prescribers to know their patients' prescription histories and accurately assess overdose risk. Wiley Periodicals, Inc.