OBJECTIVE: To assess family satisfaction with intensive care units (ICUs) in the UK using the Family Satisfaction in the Intensive Care Unit 24-item (FS-ICU-24) questionnaire, and to investigate how characteristics of patients and their family members impact on family satisfaction. DESIGN: Prospective cohort study nested within a national clinical audit database. SETTING: Stratified, random sample of 20 adult general ICUs participating in the Intensive Care National Audit & Research Centre Case Mix Programme. PARTICIPANTS: Family members of patients staying at least 24 hours in ICU were recruited between May 2013 and June 2014. INTERVENTIONS: Consenting family members were sent a postal questionnaire 3 weeks after the patient died or was discharged from ICU. Up to four family members were recruited per patient. MAIN OUTCOME MEASURES: Family satisfaction was measured using the FS-ICU-24 questionnaire. MAIN RESULTS: A total of 12 346 family members of 6380 patients were recruited and 7173 (58%) family members of 4615 patients returned a completed questionnaire. Overall and domain-specific family satisfaction scores were high (mean overall family satisfaction 80, satisfaction with care 83, satisfaction with information 76 and satisfaction with decision-making 73 out of 100) but varied significantly across adult general ICUs studied and by whether the patient survived ICU. For family members of ICU survivors, characteristics of both the family member (age, ethnicity, relationship to patient (next-of-kin and/or lived with patient) and visit frequency) and the patient (acute severity of illness and receipt of invasive mechanical ventilation) were significant determinants of family satisfaction, whereas, for family members of ICU non-survivors, only patient characteristics (age, acute severity of illness and duration of stay) were significant. CONCLUSIONS: Overall family satisfaction in UK adult general ICUs was high but varied significantly. Adjustment for differences in family member/patient characteristics is important to avoid falsely identifying ICUs as statistical outliers. TRIAL REGISTRATION NUMBER: ISRCTN47363549.