Mammography is the principal method of screening for breast cancer in the United States, however there is variability among radiologists in their interpretations of screening mammograms. This variability may cause differences in cost-effectiveness. The aim of this research was to evaluate the effect of different radiologists' practice patterns on the cost-effectiveness of screening mammography. The initial step of this analysis was an extensive literature search of published screening mammography audit results from radiology practices (five different practices were used in the final analysis). From these data, relevant parameters were selected, including screening recall rate, positive predictive value of biopsies based on mammographic findings and the Stage distribution of breast cancers diagnosed by mammography. These parameters were then inserted into a Markov model, which compares two hypothetical groups of women: the first undergoing annual mammographic screening beginning at age 40 and the second undergoing observation without screening. The model (developed using a commercial software product) is based on previously published data and calculates all of the costs and benefits accrued by each of the groups over a 39 year period. The cost-effectiveness of screening mammography (measured as marginal cost per year life saved) was calculated by comparing each of the five different practices to the observed group. The cost-effectiveness varied greatly. The most cost-effective practice had the lowest recall rate (5.20%), a high positive predictive value of biopsy (38%) and a high percentage of Stage 0 and 1 cancers (89.3%), resulting in a marginal cost per year life saved of $6,587.63 compared to the observed group. However, another practice in this study, with a recall rate of 13%, a positive predictive value of biopsy of 19% and stage 0 and 1 cancers totaling 72.1% was considerably less cost-effective. In this case, the marginal cost per year life saved of screening mammography was $21,052.67. The results suggest that the different practice patterns of individual radiologists have a profound effect on the overall cost-effectiveness of screening mammography.