Michael Rawlins continued his drive to promote the value of observational research evidence for the evaluation of medical interventions in this year’s OHE annual lecture, held at the Royal College of Physicians, London on 18th June. Citing a succession of influential thinkers and researchers throughout the history of science, he argued that the deductive approach to evidence generation, which underpins the randomised controlled clinical trial has its weaknesses as well as its well-known strengths. These include poor generalisability to the whole population who are subsequently treated outside the clinical trial setting. Observational studies in contrast, while open to bias and confounding are much more likely to provide generalisable results and a series of important observational studies including aspirin in Reye’s syndrome and ganciclovir for CMV retinitis were given as examples. He concluded that it is irrational and over-simplistic to rank RCTs above observational evidence and made a plea for decision makers to employ judgment!