Falls represent a considerable problem for many developed societies with a large growth in the elderly population, and they inflict a sizable personal and social cost. Assessment of an individual's risk of falling, facilitated by advanced movement sensing technologies, has been heralded as a potential means to target fall prevention interventions to those most at need. This paper presents the summary of a questionnaire-based survey of expert opinions collected from three major international conference meetings on the topic of fall risk assessment and fall prediction. Specifically, current problems are discussed, both in terms of technology as well as in study design, in order to harmonize and guide future research activities. The analysis of this survey confirms that sensor-based fall prediction is both relevant and promising, and that large-scale prospective studies are needed as well as further harmonization of research methods.