We have undertaken a pooled analysis of the 6 major randomized trials comparing mastectomy (MT) and breast-conserving therapy (BCT) in the treatment of primary breast cancer. Specifically, these trials compared the 2 most widely used options in local treatment: mastectomy and axillary dissection (MT) versus breast-conserving surgery, axillary dissection, and breast radiotherapy (BCT). The early results of these 6 trials formed the basis for a 1990 National Institutes of Health Consensus statement. However, most of these trials have recently published long-term follow-up results, and this pooled analysis incorporates the updated results of these 6 trials. For each of these trials, the observed number of treatment events was compared with that expected under the null hypothesis, given the number of patients per arm and the total number of events. Approximate odds ratios were computed using the observed and expected number of events, and the variance of the observed number of events. These were then pooled across trials to give overall odds ratios for the risk of locoregional recurrence, total recurrence, and death. Four of the 6 trials show that MT significantly reduces the risk of locoregional recurrence when compared with BCT, and the pooled odds ratio also shows a significant benefit for MT (odds ratio [OR], 1.561; 95% confidence interval [CI], 1.289-1.890; P < 0.001). However, only 1 trial shows a statistically significant benefit for MT in reducing mortality, and the pooled odds ratio shows no significant difference between MT and BCT (OR, 1.070; 95% CI, 0.935-1.224; P = 0.33). This pooled analysis confirms that MT and BCT have comparable effects on mortality, even after long-term follow up. However, BCT is associated with a significantly greater risk of locoregional recurrence.