The 21-gene recurrence score (Oncotype DX: RS) appears to augment clinico-pathologic prognostication and is predictive of adjuvant chemotherapy benefit in node-negative (N-) and node-positive (N+), endocrine-sensitive breast cancer. RS is a costly assay that is associated with good 'value for money' in N- disease, while economic evaluations in N+ disease based on most recent data have not been conducted. We examined the cost-utility (CU) of a RS-guided adjuvant strategy, compared to current practice without RS in N- and N+, endocrine-sensitive, breast cancer from a Canadian health care system perspective. A generic state-transition model was developed to compute cumulative costs and quality-adjusted life years (QALYs) over a 25-year horizon. Patient outcomes with and without chemotherapy in RS-untested cohorts and in those with low, intermediate and high RS were examined based on the reported prognostic and predictive impact of RS in N- and N+ disease. Chemotherapy utilization (current vs. RS-guided), unit costs and utilities were derived from a Nova Scotia Canadian population-based cohort, local unit costs and the literature. Costs and outcomes were discounted at 3% annually, and costs were reported in 2011 Canadian dollars ($). Probabilistic and one-way sensitivity analyses were conducted for key model parameters. Compared to a non-RS-guided strategy, RS-guided adjuvant therapy was associated with $2,585 and $864 incremental costs, 0.27 and 0.06 QALY gains, and resultant CUs of $9,591 and $14,844 per QALY gained for N- and N+ disease, respectively. CU estimates were robust to key model parameters, and were most sensitive to chemo utilization proportions. RS-guided adjuvant therapy appears to be a cost-effective strategy in both N- and N+, endocrine-sensitive breast cancer with resultant CU ratios well below commonly quoted thresholds.