Internationally-recognised criteria for screening for a particular disorder require the following: availability of a clear diagnosis; a suitable validated screening test; acceptability of routine screening by patients and health providers; benefits of earlier detection and application of appropriate interventions to prevent the progression of a disorder, and hence reduction in incidence of morbidity and mortality; identification of possible harm from screening (false positives, false negatives, adverse effects of labelling, early diagnosis or unnecessary treatment of persons with true-positive test results with inconsequential disease) and weighing this against potential benefits; identification of possible sub-populations with the possibility of targeted screening of high-risk populations; good quality evidence of interventions effective in preventing or managing the disorder; and a cost-effectiveness assay. The New Zealand Ministry of Health have launched a best-practice guideline recommending all female general practice patients sixteen years and over be routinely screened for physical and sexual abuse by their partners. Inter-partner violence, especially against women by male partners and expartners, is a serious public health problem. However, review of existing research indicates that this guideline meets none of the criteria listed above. Considerable funding is invested in training health providers to implement this screening protocol, but, in the absence of effectiveness studies, cost-effectiveness cannot be assessed. Under current conditions, routine screening of adult women for partner abuse cannot be justified. However, GPs should be encouraged to learn about partner abuse and consider this possibility in patients presenting with physical injuries, psychological disturbance or social dysfunction, especially in high-risk patients. Research should be supported for the development and validation of effective, acceptable screening tools and randomised controlled trials of appropriate interventions. The desire to intervene for the public good should not dictate the implementation of a screening programme that disregards accepted screening criteria.