BACKGROUND: Older crime victims may be particularly vulnerable to psychological distress. OBJECTIVES: To compare the clinical effectiveness of a Victim Improvement Package (VIP) to treatment as usual (TAU) for reducing continued crime-associated distress. DESIGN: A three-step parallel-group single-blind randomised controlled trial. SETTING: Police-reported victims of community crime aged 65 and over were recruited from 12 local authority areas in a major urban city in England, UK. PARTICIPANTS: Selection criteria-inclusion: victims of community crime aged 65 years or more, with significant Generalised Anxiety Disorder (GAD-2) and Patient Health Questionnaire (PHQ-2) distress associated with the crime. EXCLUSION: type of crime, diagnosis, receipt of cognitive-behavioural therapy (CBT) in the last 6 months; an inability to participate in CBT; cognitive impairment. Participants were typical of our local authority population; 71% were female, 69% white, with the majority of crimes associated with burglary (35%) and theft (26%). 67% (88/131) of the randomised participants were included in the primary analysis. INTERVENTIONS: TAU was compared with TAU plus up to 10 sessions of a cognitively-behaviourally informed VIP, delivered by a mental health charity over 12 weeks. PRIMARY AND SECONDARY OUTCOME MEASURES: Timings are in relation to the crime; baseline (3 months), post-VIP intervention (6 months) and follow-up (9 months). The primary outcome was a composite of the Beck Anxiety and Beck Depression Inventories. The primary endpoint was 6 months. RESULTS: 24% (4255/17 611) of reported crime victims were screened, 35% (1505/4255) were distressed. Of 60% (877/1505) rescreened at 3 months, 49% (427/877) remained distressed. Out of our target of 226, 131 participants were randomised; 65 to VIP and 66 to TAU alone. 68% (89/131) completed the primary outcome (post-intervention). The VIP showed no overall benefit; mean VIP -0.41 (SD 0.89) vs mean TAU -0.19 (SD 1.11); standardised mean difference -0.039; 95% CI (-0.39, 0.31), although stratified analyses suggested an effect in burglary victims (n=27, standardised mean difference -0.61; 95% CI (-1.22, -0.002), p=0.049). CONCLUSIONS: Community crime had long-lasting impacts. The police are ideally placed to screen for distress, present in 35% of victims, but only 58% of participants were recruited and the cognitive-behavioural therapy was not delivered competently. Further research on victim care and improving the delivery and quality of therapy is required. TRIAL REGISTRATION NUMBER: All procedures were approved by the University College London (UCL) Research Ethics Committee on 17 March 2016 (6960/001). International Standard Randomised Controlled Trial Number is ISRCTN16929670: https://doi.org/10.1186/ISRCTN16929670.