Background During COVID-19 pandemic, visits have been prohibited in most French ICUs. Psychological effects, for reference persons (RPs), of remote-only communication have been assessed. Methods All RPs of patients referred to ICU for COVID-19 were included. HADS, IES-R, and satisfaction were evaluated at admission, discharge/death, and 3 months. At 3 months, a psychologist provided a qualitative description of RPs’ psychological distress. Results Eighty-eight RPs were included. Prevalence of anxiety and depression was 83% and 73% respectively. At 3 months, lower HADS decrease was associated with patient death/continued hospitalization, and/or sleeping disorders in RPs (p < 0.01). Ninety-nine percent RPs felt the patient was safe (9 [7; 10]/10 points, Likert-type scale), confident with caregivers (10 [9; 10]/10 points), and satisfied with information provided (10 [9; 10]/10 points). All RPs stressed the specific-type of “responsibility” associated with being an RP in a remote-only context, leading RPs to develop narrow diffusion strategies (67%) and restrict the array of contacted relatives to a very few and/or only contacting them rarely. 10 RPs (30%) related the situation to a prior traumatic experience. Conclusion RPs experienced psychological distress and reported that being an RP in a remote-only communication context was a specific responsibility and qualified it as an overall negative experience. Trial registration NCT04385121. Registered 12 May 2020. https://clinicaltrials.gov/.