OBJECTIVES: Meta-analysis of the effect of mammography on breast cancer (BC) mortality, using random effects (RE) inverse-variance weighted techniques, accounting for heterogeneity by incorporating a quality weight into the analysis. METHODS: Raw data were extracted from the most relevant publications of eight RCTs of mammography screening identified through a systematic review. Quality-weighted random effects meta-analysis (QWREMA) estimated the pooled relative risk (RR) of BC mortality in screened vs. control groups, in women under 50, over 50, and of all ages by empirically weighting weighting the between-study variance with a quality score assigned to each trial derived through qualitative review of its supportive literature. The mammography-specific quality rating system contained both scale elements and a component approach. Results were compared to those from fixed effects (FE) andRE methods. Sensitivity analysis assessed the impact of inclusion criteria on the results andcumulative analysis assessed the effect of decreasing order of quality. RESULTS: A significant 20% reduction in BC mortality was observed in women of any age [RR: 0.80(95%CI=0.72-0.87)], and 22% reduction in women over 50 [RR: 0.78(95%CI=0.70-0.88)]. Lack of between-study variability in women under 50 reduced the result to a FE analysis: 0.84(95%CI=0.73-0.97). Results were robust to different weights on algorithm quality domains. QWREMA tightened the RE 95%CI by 10% in the >50 group, and by almost 12% in women of all ages. No specific trend of effect with study quality was observed in women of all ages and women >50. In women <50, 7 studies had to be pooled before a statistically significant beneficial effect was observed. CONCLUSION: Quality scores may be used in a meta-analysis to account for between-study variance due to heterogeneity. The efficacy of mammography in the 40-49 age-group remains in question, and may depend on the number and/or quality of studies included in the meta-analysis.