Background: Pain management can be challenging in frail older people with cancer due to drug–drug interactions and heightened susceptibility to adverse drug events. Objective: To investigate the relationship between analgesic use and pain by frailty status in older outpatients with cancer. Methods: A total of 385 consecutive patients aged 70 years and over who presented to an outpatient oncology clinic between January 2009 and July 2010 completed structured assessments of analgesic use (opioids, paracetamol or non-steroidal anti-inflammatory drugs), pain (10-point visual analogue scale) and clinical factors. Frailty was derived using modified Fried’s frailty phenotype. Logistic regression was used to compute adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) for the relationship between analgesic use and pain for each frailty group (robust, pre-frail or frail). Results: For robust outpatients (n = 101), there was weak evidence for a 30 % relative increase in the adjusted odds of analgesic use between outpatients who differed by one unit of pain score (95 % CI 0.995−1.71, p = 0.0532). For pre-frail outpatients (n = 190), there was evidence for a negative quadratic relationship (adjusted OR for the quadratic coefficient: 0.952, 95 % CI 0.910−0.993, p = 0.0244). For frail outpatients (n = 94), there was an 8 % relative increase in the adjusted odds of analgesic use between outpatients who differed by one unit of pain score, but no statistical evidence for association (95 % CI 0.934−1.26; p = 0.298). Conclusions: These findings can be considered for the ongoing development of safe, effective strategies for analgesic use in older outpatients with cancer.