Background Although surgical resection has been considered the only curative option for colorectal liver metastases, thermal ablation has recently been suggested as an alternative curative treatment. There have been no adequately powered trials comparing surgery with thermal ablation. Objectives Main objective – to compare the clinical effectiveness and cost-effectiveness of thermal ablation versus liver resection surgery in high surgical risk patients who would be eligible for liver resection. Pilot study objectives – to assess the feasibility of recruitment (through qualitative study), to assess the quality of ablations and liver resection surgery to determine acceptable standards for the main trial and to centrally review the reporting of computed tomography scan findings relating to ablation and outcomes and recurrence rate in both arms. Design A prospective, international (UK and the Netherlands), multicentre, open, pragmatic, parallel-group, randomised controlled non-inferiority trial with a 1-year internal pilot study. Setting Tertiary liver, pancreatic and gallbladder (hepatopancreatobiliary) centres in the UK and the Netherlands. Participants Adults with a specialist multidisciplinary team diagnosis of colorectal liver metastases who are at high surgical risk because of their age, comorbidities or tumour burden and who would be suitable for liver resection or thermal ablation. Interventions Thermal ablation conducted as per local policy (but centres were encouraged to recruit within Cardiovascular and Interventional Radiological Society of Europe guidelines) versus surgical liver resection performed as per centre protocol. Main outcome measures Pilot study – patients’ and clinicians’ acceptability of the trial to assist in optimisation of recruitment. Primary outcome – disease-free survival at 2 years post randomisation. Secondary outcomes – overall survival, timing and site of recurrence, additional therapy after treatment failure, quality of life, complications, length of hospital stay, costs, trial acceptability, and disease-free survival measured from end of intervention. It was planned that 5-year survival data would be documented through record linkage. Randomisation was performed by minimisation incorporating a random element, and this was a non-blinded study. Results In the pilot study over 1 year, a total of 366 patients with colorectal liver metastases were screened and 59 were considered eligible. Only nine participants were randomised. The trial was stopped early and none of the planned statistical analyses was performed. The key issues inhibiting recruitment included fewer than anticipated patients eligible for both treatments, misconceptions about the eligibility criteria for the trial, surgeons’ preference for one of the treatments (‘lack of clinical equipoise’ among some of the surgeons in the centre) with unconscious bias towards surgery, patients’ preference for one of the treatments, and lack of dedicated research nurses for the trial. Conclusions Recruitment feasibility was not demonstrated during the pilot stage of the trial; therefore, the trial closed early. In future, comparisons involving two very different treatments may benefit from an initial feasibility study or a longer period of internal pilot study to resolve these difficulties. Sufficient time should be allowed to set up arrangements through National Institute for Health Research (NIHR) Research Networks. Trial registration Current Controlled Trials ISRCTN52040363. Funding This project was funded by the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 21. See the NIHR Journals Library website for further project information.