Annual surveillance mammography is commonly recommended after breast conservation therapy (BCT). We retrospectively reviewed its effectiveness on 511 invasive and non-invasive breast cancers treated with BCT between 1994 and 2003. The median follow-up was 5.9 years. The 5-year actuarial ipsilateral breast tumour recurrence (IBTR) rate was 4.5% and contralateral breast cancer (CBC) rate was 2.0% (representing eight times increase in risk). IBTR of 43% and 62% CBC were first detected by surveillance mammography. The IBTR detection rates per 1000 mammograms were 5.2 for patients (n=349) with one or more IBTR risk factors (age 45, positive/close margins or histological grade 3) and 0.6 for patients (n=162) without. No survival difference was observed between different modes of IBTR detection (p=0.342). In conclusion, a risk-adapted approach of limiting ipsilateral surveillance to patients with IBTR risk is possible but its implementation will be complicated by the continued need of contralateral surveillance.