Background Protein energy malnutrition (PEM) predisposes individuals to disease, delays recovery from illness and reduces quality of life. Care home residents in the United Kingdom are especially vulnerable, with an estimated 30 to 42 % at risk. Evidence for nutritional interventions to address PEM in the care home setting is lacking. Widely used techniques include food-based intervention and/or the use of prescribed oral nutritional supplements. To define outcomes and optimise the design for an adequately powered definitive trial to compare the efficacy of established nutritional interventions in this setting, a cluster randomised feasibility trial with a 6-month intervention was undertaken. Methods Care home residents with or at risk of malnutrition were identified across six UK care home sites from September to December 2013. Homes were cluster randomised to standard care (SC), food-based intervention (FB) or oral nutritional supplement intervention (ONS), for 6 months. Key outcomes were trial feasibility and the acceptability of design, allocated interventions and outcome assessments. Anthropometry, dietary intake, healthcare resource usage and participant-reported outcome measures were assessed at baseline and at 3 and 6 months. Results All six care homes approached were recruited and retained. Of the 110 residents at risk of malnutrition, 85 % entered the trial, and 68 % completed the 6-month intervention. Pre-specified success criteria for feasibility were met for recruitment and retention, intervention acceptability (resident compliance ≥60 %) and measurement of weight, body mass index (BMI), mid-upper arm circumference and dietary intake (data completeness >80 %). Measurement of handgrip strength and triceps skinfold thickness was not found to be feasible in this population. The 95 % confidence interval (CI) data suggested sensitivity to change in dietary intake for weight, BMI and energy intake between baseline and 3 months when each intervention (FB and ONS) was compared with SC. Conclusions A definitive trial comparing the efficacy of nutritional support interventions in increasing weight and BMI in malnourished care home residents can be conducted. However, whilst the design was feasible, this trial has highlighted the lack of clinically and patient-relevant outcome measures that are appropriate for use in this setting for both research and clinical practice. In particular, this trial identified a need for a more simple measure of functional status, which considers the limitations of functional tests in the care home population.