Aim To describe short‐term mortality among residential aged care (RAC) residents in Auckland, New Zealand. Method Prospective follow‐up, 6828 residents (median age 86 years, 69.8% women) from census‐type survey (10/9/2008); 152 facilities. Mortality data from central sources. Results Eight hundred and sixty‐one (12.6%) died by 6 months. Survival related to RAC length of stay before the survey: those resident <1 month (subgroup, n = 380) having 80.0% survival, 1–6 months 83.2% and >6 months 87.4% (P < 0.0001). In those admitted to private hospital from acute hospital (n = 104 of the subgroup of 380), 6‐month mortality was 36.5% (P < 0.0001 vs other ‘short stayers’). Significant mortality predictors were: private hospital admission from acute hospital (hazard ratio (HR) = 2.02), unscheduled GP visit during the prior 2 weeks (HR= 1.90), personal care disability (HR= 1.90) and acute hospital admission number during previous 2 years (≥3; HR= 5.40). Conclusions RAC mortality (especially post admission) is high. Training and resource in the sector should reflect this.