Background: Unintentional injury is the leading cause of preventable death of children over the age of 1 year in the UK and a major cause of attendance at emergency departments. Children having one injury are at increased risk of further injuries. Parenting programmes can reduce injuries in preschool children if delivered in the home and on a one-to-one basis. It is not known if group-based programmes delivered outside the home are effective. Objectives: To develop (1) a parenting programme to prevent recurrent unintentional home injuries in preschool children and (2) a tool for parents to report unintentional home injuries occurring to their preschool children. To assess the feasibility of delivering and evaluating the parenting programme through a cluster randomised controlled trial, specifically to (1) assess methods for the recruitment and retention of parents; (2) determine the training, equipment and facilities needed for the delivery of the programme; (3) establish appropriate primary and secondary outcome measures and methods for their collection; (4) determine how 'normal care' in a comparison arm should be defined; and (5) determine the resource utilisation and costing data that would need to be collected for the cost-effectiveness component of a future trial; and (6) produce estimates of effect sizes to inform sample size estimation for a main trial. Design: Feasibility multicentre, cluster, randomised, unblinded trial. Setting: Eight children's centres in Bristol and Nottingham, UK. Participants: Ninety-six parents of preschool children who had sustained an unintentional injury requiring medical attention in the previous 12 months. Interventions: The First-aid Advice and Safety Training (FAST) parent programme, comprising parenting support and skills combined with first aid and home safety advice. Main outcome measures: Parent-reported medically attended injuries in the index child and any preschool siblings sustained during a 6-month period of observation. Results: An 8-week parenting programme was produced, designed with participant-friendly, incrementally progressive content. A slimline, month-to-a-view injury calendar, spiral bound and suitable for hanging on a wall, was designed for parents to record injuries occurring to their preschool children during the 6-month period of observed time. Fifty-one parents were recruited (40 meeting eligibility criteria plus 11 following 'open invite' to participate); 15 parents completed the FAST parent programme and 49 provided data at baseline and during follow-up. Completion of the programme was significantly greater for participants using the 'open invite' approach (85%) than for those recruited using the original eligibility criteria (31%). Prototype resource use checklists, unit costs and total costs were developed for phases 0, 1 and 2 of the study for use in a future trial. Conclusions: This feasibility study has developed an innovative injury prevention intervention and a tool to record parent-reported injuries in preschool children. It was not feasible to recruit parents of children who had sustained a recent injury, or to ask health visitor teams to identify potential participants and to deliver the programme. A trial should target all families attending children's centres in disadvantaged areas. The intervention could be delivered by a health professional supported by a member of the children's centre team in a community setting. Trial registration: Current Controlled Trials ISRCTN03605270. Source of funding: This project was funded by the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 18, No. 3. See the NIHR Journals Library website for further project information. © Queen's Printer and Controller of HMSO 2014.