BackgroundOver one in five UK doctors reports feeling burnt out because of their work. This study aimed to determine if a novel text message intervention could reduce burnout in UK trainee anaesthetists. MethodsIn this two-arm, non-blinded, randomised controlled trial, we recruited 2nd to 4th year trainees registered with the Royal College of Anaesthetists, UK. Eligible trainees were given information about the trial and asked to provide their mobile phone numbers to indicate consent to participate. Participation was optional. Participants who signed up to the trial were randomly assigned (using a random number generator in the statistical programme Stata) to receive either fortnightly text messages over 10 months (treatment group) or no intervention messages (control group). Randomisation was stratified by training year and region. A literature review and qualitative study informed the intervention design and content: our intervention was low cost to deliver and minimally time-intensive for participants (contrasting many previous interventions), and it aimed to address factors identified as contributing to burnout. The message content also drew on 11 themes from the behavioural science and wellbeing literature (eg, gratitude, social support, self-compassion) to improve burnout and wellbeing. The co-primary outcomes were burnout (measured by the work-related subscale of the Copenhagen Burnout Inventory) and wellbeing (Short Warwick-Edinburgh Mental Wellbeing Scale). Secondary outcomes were meaning in work, professional value, sickness absence, and consideration of career break. Exploratory outcomes were factors potentially affecting wellbeing, including COVID-19. Outcomes were measured via an online survey. Intervention impact for those who completed the endline survey was evaluated using regression adjusting for gender, training year, training region, and baseline measurement. We conducted exploratory analyses of intervention impact in subgroups of participants using the same regression model. The Imperial College Research Ethics Committee approved the study on Aug 14, 2019 (reference number 19IC5205). The Imperial Joint Research Compliance Office approved the additional exploratory outcomes in an amendment on July 21, 2020. This trial is registered with the ISRCTN registry, ISRCTN11418903. FindingsIn September-October 2019, 279 trainees were randomised, of whom 153 (80 female; 73 male) completed the endline survey (74 in the treatment group; 79 in the control group). There were no significant group differences in burnout (β=–1·82, 95% CI –6·54 to 2·91, p=0·45), wellbeing (–0·52, –1·73 to 0·69, p=0·40), or secondary outcomes (all p values ≥0·24). However, 49 (66%) of 74 participants in the treatment group recommended the messages be sent to future trainees. Exploratory analyses found the intervention was associated with reduced burnout in participants who reported personal or work-related difficulties (β=–9·56, –17·35 to –1·77), and those for whom COVID-19 had a large negative wellbeing impact (β=–10·38, –20·57 to –0·19). InterpretationThe intervention had no impact overall, but was recommended by most, and associated with reduced burnout in participants reporting difficult personal or work circumstances. Given this intervention is low cost and requires minimal time commitment, it may warrant adaptation (eg, based on participant feedback) and further evaluation. FundingNational Institute for Health Research.