Debate is ongoing about the nature and use of evidence in public health decision making, and there seems to be an emerging consensus that the "hierarchy of evidence" may be difficult to apply in other settings. It may be unhelpful however to simply abandon the hierarchy without having a framework or guide to replace it. One such framework is discussed. This is based around a matrix, and emphasises the need to match research questions to specific types of research. This emphasis on methodological appropriateness, and on typologies rather than hierarchies of evidence may be helpful in organising and appraising public health evidence.