The objective of this project was to analyze newspaper coverage of the January 2000 meta-analysis by Gotzsche and Olsen, "Is screening for breast cancer with mammography justified?" [Lancet 355 (2000) 129]. A content analysis was performed on a comprehensive set of newspaper clippings from the UK during the 2 weeks following publication of the Lancet article. The original authors were most quoted in Wave 1 (the first weekend); the screening programme was most quoted in Wave 2 (week 2). Screening programme description, and the "quality" of the Lancet article dominated Wave 1; patient testimonials increased in Wave 2. Newspaper articles were structured as debates between experts and advocates, thereby enhancing polarisation of opinion. We suggest this is counter-productive to evidence-based patient choice and public involvement in decision-making. Medical journals' and charities' press releases that begin to include discussion of uncertainty inherent in medical technologies can contribute to evidence-based public deliberation.