This paper describes the development and testing of a lameness control programme (LCP) for heifers on dairy farms. The LCP, which is based on the analysis of hazards and critical control points (HACCP), was tested via a randomised intervention study on 60 farms. Tangible hazards for each farm were identified, allotted to 11 categories of proximate hazard and scored on each farm to quantify the risks presented by each hazard. Feet were inspected for signs of claw horn disease and infection, such as digital dermatitis. Intervention was generally ineffective, primarily through failure to implement the LCP. However, retrospective analysis was able to demonstrate highly significant associations between risks attached to proximate hazards and probabilities of lameness and foot lesions, allowing the severity of these hazards to be ranked. The most significant proximate hazards of environmental origin were prolonged standing on concrete, standing in wet slurry and factors that cause claw trauma. The most severe proximate hazards however were those associated with failures of management, especially poor claw condition and inadequate foot care. Overall farm risks (OFR) were estimated by summing the products of the generic severity for proximate hazards with on-farm risks. Changes in OFR were significantly related to changes in outcome (lameness and lesions).