The debate about how best to determine nurse staffing levels continues. The conventional wisdom is that determining staffing levels is something best left to local management, taking account of local workload and resources. This 'bottom up' philosophy has now been challenged by the use of a different approach--the use of 'top down'standardized, and mandatory, nurse:patient or nurse:bed ratios. This paper examines the characteristics and early results of the use of staffing ratios in the two health systems where nurse staffing ratios are now mandatory--the states of Victoria (Australia) and California (USA). It then discusses the policy implications of using ratios. The paper identifies the main weaknesses of the use of nurse:patient ratios as being their relative inflexibility and their potential inefficiency, if they are wrongly calibrated. Their strength is their simplicity and their transparency. Their impact will be most pronounced when ratios are mandatory and where they offer a mechanism to improve and then to maintain staffing levels at some pre-determined level. The biggest challenges in their use are calibration (what is 'safe'? or 'minimum'?) and achieving the support of all stake-holders. The paper concludes that nurse:patient ratios are a blunt instrument for achieving employer compliance, where reliance on alternative, voluntary (and often more sophisticated) methods of determining nurse staffing have not been effective.