- BACKGROUND: Basal-cell carcinoma is the most common form of skin cancer and its incidence is increasing worldwide. We aimed to assess the effectiveness of imiquimod cream versus surgical excision in patients with low-risk basal-cell carcinoma. METHODS: We did a multicentre, parallel-group, pragmatic, non-inferiority, randomised controlled trial at 12 centres in the UK, in which patients were recruited between June 19, 2003, and Feb 22, 2007, with 3 year follow-up from June 26, 2006, to May 26, 2010. Participants of any age were eligible if they had histologically confirmed primary nodular or superficial basal-cell carcinoma at low-risk sites. We excluded patients with morphoeic or recurrent basal-cell carcinoma and those with Gorlin syndrome. Participants were randomly assigned (1:1) via computer-generated blocked randomisation, stratified by centre and tumour type, to receive either imiquimod 5% cream once daily for 6 weeks (superficial) or 12 weeks (nodular), or surgical excision with a 4 mm margin. The randomisation sequence was concealed from study investigators. Because of the nature of the interventions, masking of participants was not possible and masking of outcome assessors was only partly possible. The trial statistician was masked to allocation until all analyses had been done. The primary outcome was the proportion of participants with clinical success, defined as absence of initial treatment failure or signs of recurrence at 3 years from start of treatment. We used a prespecified non-inferiority margin of a relative risk (RR) of 0.87. Analysis was by a modified intention-to-treat population and per protocol. This study is registered as an International Standard Randomised Controlled Trial (ISRCTN48755084), and with ClinicalTrials.gov, number NCT00066872. FINDINGS: 501 participants were randomly assigned to the imiquimod group (n=254) or the surgical excision group (n=247). At year 3, 401 (80%) patients were included in the modified intention-to-treat group. At 3 years, 178 (84%) of 213 participants in the imiquimod group were treated successfully compared with 185 (98%) of 188 participants in the surgery group (RR 0.84, 98% CI 0.78-0.91; p<0.0001). No clear difference was noted between groups in patient-assessed cosmetic outcomes. The most common adverse events were itching (211 patients in the imiquimod group vs 129 in the surgery group) and weeping (160 vs 81). We recorded serious adverse events in 99 (40%) of 249 participants in the imiquimod group and 97 (42%) of 229 in the surgery group had serious adverse events, but none were regarded as related to treatment. 12 (5%) participants in the imiquimod group withdrew because of adverse events compared with four (2%) in the surgery group. INTERPRETATION: Imiquimod was inferior to surgery according to our predefined non-inferiority criterion. Although excisional surgery remains the best treatment for low-risk basal-cell carcinoma, imiquimod cream might still be a useful treatment option for small low-risk superficial or nodular basal-cell carcinoma dependent on factors such as patient preference, size and site of the lesion, and whether the patient has more than one lesion.