On behalf of certain Unilever marketing companies, Research Bureau conducts continuous brand image measurements under the title of the "Brand Performance Survey". This survey involves, for a given product field, some 6,000 interviews in a year with a multi-stage national sample of urban housewives; its quarterly reports contain data on awareness of, and attitudes to each major brand and its advertising. Although primarily designed to replace, by economic continuous research, ad hoc image studies, the BPS naturally includes specific measures of the type commonly employed in the assessment of the effect of advertising campaigns. During 1965, the Brand Performance Survey was used as a vehicle for investigating the apparent relationships of these measures to purchasing behaviour. In July of that year, re-contact was attempted with 4, 500 respondents first interviewed between July, 1964 and March, 1965; a success rate of approximately 75% was achieved. Four product fields were covered. At the second interview, the usership questions forming part of the regular BPS questionnaire were repeated. The present paper is restricted to results relating to one field only, the product being a low-ticket, frequently purchased item with high penetration. For simplicity, the data to be cited is only for the shortest period between initial interview and recontact, averaging five months approximately.