The research question is a crucial aspect of every study. Criteria for the evaluation of the merit of the study question or hypothesis have received surprisingly little attention. A set of non-methodological criteria derived from interviews with 40 researchers (clinical investigators and laboratory scientists) is presented. For both types of researcher, the terms that best described the nature of the criteria included potential impact, justification, feasibility, track record, innovation, intuitive response, aesthetics and politics. The latter three criteria are notably subjective; however all the criteria have an element of non-objectivity. There may well be a reluctance to openly acknowledge that crucial choices about what scientific questions should be explored involve criteria which are themselves not 'objective', and indeed, not even 'scientific'.