BACKGROUND: Recruitment to randomised controlled trials (RCTs) with very different treatment arms is often difficult. The ProtecT (Prostate testing for cancer and Treatment) study successfully used qualitative research methods to improve recruitment and these methods were replicated in five other RCTs facing recruitment difficulties. A similar qualitative recruitment investigation was undertaken in the SPARE (Selective bladder Preservation Against Radical Excision) feasibility study to explore reasons for low recruitment and attempt to improve recruitment rates by implementing changes suggested by qualitative findings. METHODS: In Phase I of the investigation, reasons for low levels of recruitment were explored through content analysis of RCT documents, thematic analysis of interviews with trial staff and recruiters, and conversation analysis of audio-recordings of recruitment appointments. Findings were presented to the trial management group and a plan of action was agreed. In Phase II, changes to design and conduct were implemented, with training and feedback provided for recruitment staff. RESULTS: Five key challenges to trial recruitment were identified in Phase I: (a) Investigators and recruiters had considerable difficulty articulating the trial design in simple terms; (b) The recruitment pathway was complicated, involving staff across different specialties/centres and communication often broke down; (c) Recruiters inadvertently used 'loaded' terminology such as 'gold standard' in study information, leading to unbalanced presentation; (d) Fewer eligible patients were identified than had been anticipated; (e) Strong treatment preferences were expressed by potential participants and trial staff in some centres. In Phase II, study information (patient information sheet and flowchart) was simplified, the recruitment pathway was focused around lead recruiters, and training sessions and 'tips' were provided for recruiters. Issues of patient eligibility were insurmountable, however, and the independent Trial Steering Committee advised closure of the SPARE trial in February 2010. CONCLUSIONS: The qualitative investigation identified the key aspects of trial design and conduct that were hindering recruitment, and a plan of action that was acceptable to trial investigators and recruiters was implemented. Qualitative investigations can thus be used to elucidate challenges to recruitment in trials with very different treatment arms, but require sufficient time to be undertaken successfully. TRIAL REGISTRATION: CRUK/07/011; ISRCTN61126465.