Exposure to opioid analgesics due to surgery increases the risk of new persistent opioid use. A mechanistic hypothesis for opioids' abuse liability rests on the belief that, in addition to pain relief, acute opioid treatment improves well-being (e.g. via euphoria) and relieves anxiety. However, opioids do not consistently improve mood in laboratory studies of healthy non-opioid users. This observational study determined how two commonly used opioid analgesics affected patients' subjective well-being in standard clinical practice. Day surgery patients rated how good and how anxious they felt before and after an open-label infusion of remifentanil (n = 159) or oxycodone (n = 110) in the operating theatre before general anaesthesia. One minute after drug injection, patients reported feeling intoxicated (> 6/10 points). Anxiety was reduced after opioids, but this anxiolytic effect was modest (remifentanil Cohen's d = 0.21; oxycodone d = 0.31). There was moderate to strong evidence against a concurrent improvement in well-being (Bayes factors > 6). After remifentanil, ratings of 'feeling good' were significantly reduced from pre-drug ratings (d = 0.28). After oxycodone, one in three participants felt better than pre-drug. Exploratory ordered logistic regressions revealed a link between previous opioid exposure and opioid effects on well-being, as only 14 of the 80 opioid-naïve patients reported feeling better after opioid injection. The odds of improved well-being ratings after opioids were higher in patients with previous opioid exposure and highest in patients with > 2 weeks previous opioid use (adjusted OR = 4.4). These data suggest that opioid-induced improvement of well-being is infrequent in opioid-naïve patients. We speculate that peri-operative exposure could increase risk of persistent use by rendering subsequent positive opioid effects on well-being more likely.