Purpose: To determine the effect of introducing radiographer double reading, in addition to standard radiologist double reading, on screening mammography outcome. Methods: In period A, 66,225 mammograms were read by two screening radiologists. In period B, 78,325 mammograms were read by two radiographers in addition and radiologists were blinded to the referral opinion of the radiographers. Mammograms, for which only radiographers had suggested referral, (i.e. cases that would only be referred by technologists) were re-evaluated by the screening radiologists. Women were referred if at least one radiologist considered this necessary, and diagnostic costs of these additional referrals were estimated. Results: In period A, 322 cancers were diagnosed after referral of 678 women. During period B, radiologists initially referred 1122 patients and 411 cancers were detected. Radiologists' referral rate was higher in period B than in period A (1.43% versus 1.02%, p < 0.001), as well as the cancer detection rate per 1000 women screened (CDR) (5.25 versus 4.86, p = 0.3). The positive predictive value of referral (PPV) was 36.6% versus 47.5% (p < 0.001). In period B, radiologist review of 544 additional positive radiographer readings led to 102 extra referrals, with 29 additional cancers detected, resulting in an overall referral rate of 1.56% (compared to period A, p < 0.001), an overall CDR of 5.62 (p = 0.048) and an overall PPV of 35.9% (p < 0.001). Workup expenses of the 102 additional referrals were Sic60,274. Conclusion: Additional radiographer double reading detected cancers that would have been missed by radiologists. Mean expenses for diagnostic confirmation of these extra cancers was Sic2078 per cancer. (C) 2008 Elsevier Ltd. All rights reserved.