Late cardiac toxicity is a substantial concern in women receiving radiation therapy for left-sided breast cancer. The 2005 Oxford overview identified an increased risk of cardiac mortality in this cohort; however these data represent patients treated in the 2D era. We sought to characterize heart dose in patients with left-sided breast cancer treated with adjuvant radiation spanning an eleven-year period, representing the earliest patients to undergo three-dimensional planning at our institution through the present era. Plans of 310 patients with left-sided breast cancer between 2003 and 2013 were reviewed. All patients received supine whole breast or chest wall radiation with standard fractionation (45-50 Gy). We recorded mean heart dose and V25 Gy and V5 Gy for the whole heart (WH), left ventricle (LV), right ventricle (RV) and both ventricles (BV). Patients were separated into three periods: 2003-2004 (when three-dimensional planning was first instituted, n = 29), 2005-2006 (following publication of the Oxford overview, n = 22), and 2009-2013 (n = 259). Patients in the two earlier groups did not have heart contours or DVHs assessed at the time of planning. For these patients we retrospectively contoured the heart volumes and assessed the dose delivered to these structures in the initial plans. ANOVA methods for repeated measures were applied to investigate the difference in mean values over time, with subsequent Newman-Keuls pairwise post hoc tests used for pairwise comparisons if the overall test was significant (p < 0.05). Mean heart dose was 345 cGy (2003-4; range 114-693 cGy), 212 cGy (2005-6; range 133-499 cGy), and 213 cGy (2009-13; range 91-445 cGy). LV V25 Gy was 6.3% (2003-4), 1.4% (2005-6), and 1.1% (2009-13). LV V5 Gy was 15.2% (2003-4), 8% (2004-5), and 6.8% (2009-13). RV V25 Gy was 3.3% (2003-4), 0.5% (2004-5), and 0.2% (2009-13). RV V5 Gy was 12.1% (2003-4), 8% (2004-5), and 3.2% (2009-13). BV V25 Gy was 5.1% (2003-4), 1.1% (2004-5), and 0.9% (2009-13). BV V5 Gy was 13.9% (2003-4), 6.4% (2004-5), and 5.5% (2009-13). For each model ANOVA overall results indicated significant differences among means for all parameters (p < 0.0001 for each). Post hoc tests showed significantly higher doses in patients treated during 2003-2004 compared to those treated in each of the later year cohorts (p < 0.002) for all parameters except RV V25 Gy (p = 0.23). Heart doses were significantly higher for women treated during the earliest years of three-dimensional planning, compared to subsequent years. Heightened awareness of radiation cardiac toxicity after 2005 related to major publications on this topic may have influenced this trend. Data from the earliest cohort, in which heart volumes were not drawn and DVHs were not calculated, give some insight into the heart doses that may have been delivered prior to the 3D era.