The Hamilton Depression Rating Scale, Beck Depression Inventory, Clinical Interview Schedule and General Health Questionnaire have been designed to detect and measure severity of psychiatric disorders in different settings. We aimed to determine the relationship between each assessment within a mixed primary care population. The questionnaires and interviews were administered to a sample of 49 patients attending an urban general practice. The total score for each scale correlated closely with every other one (correlation coefficients = 0.70-0.79). Factor analysis suggested that all of the scales were measuring a single underlying construct with reliabilities of 0.68-0.78. The implications of these findings are discussed in relation to the assessment of severity of depression and other neurotic symptoms. We suggest that the Clinical Interview Schedule is a feasible alternative to the Hamilton Rating Scale for Depression and discuss reasons why it might be preferable within clinical trials of depression.