Various dairy cattle production and health characteristics were studied with a view to identify easily available and measurable factors associated with the incidence of ketosis. The analyses were carried out using data from the Langhill Dairy Cattle Research Centre. Two approaches were used to assess the relative risk to cows of getting ketosis: one using information known at the beginning of lactation and one using information collected as the lactation progressed. In both approaches analyses were carried out using different amounts of the available information to simulate differences between recording systems. In the first approach the following were found to relate significantly to the level of recorded ketotic incidents: parity; ketosis in the previous lactation; calving condition score; 305-day milk yield in the previous lactation; and the average milk protein percentage in the previous lactation. The effects of these were quantified. In the second approach, where the change in ketosis incidence rate over the weeks of lactation was investigated, the average dry-matter intake in the previous week and changes in live weight and body condition score over the previous week were found to have a significant effect on the probability of getting ketosis in the coming week. The risk assessments varied depending on the information used and a flexible approach is recommended if potential risk factors are to be successfully incorporated into decision support systems.