Aim: Existing studies on the association between haemodialysis facility size/volume and patient survival are mostly limited to freestanding dialysis units in the United States. This study in Taiwan explored the facility size – mortality association in both hospital-based and freestanding haemodialysis (HD) units. Methods: In this nationwide population-based retrospective cohort study, we used the Taiwan National Health Insurance Research Database to include patients who began maintenance (HD) between 2008 and 2012. Facility size was categorized according to the number of stations in the HD unit. The 5 years mortality rate was analyzed using a frailty model for Cox regression. The patients in hospital-based and freestanding HD units were examined separately. Results: Among the 39 506 patients, 24 597 (62.3%) and 14 909 (37.7%) patients received HD in hospital-based and freestanding facilities, respectively. After the 4th month of dialysis initiation, the 5 years survival rates of patients in hospital-based and freestanding HD units were 50.7% and 52.3%, respectively. When patient and other facility characteristics were adjusted, patients in the smallest facility category (1–15 stations) showed the highest mortality risk (hazard ratio, 1.36; 95% confidence interval, 1.11–1.67) among all the patients treated in hospital-based units. The patients treated in freestanding units with 1–15, 16–30 and 31–45 stations showed 31%, 33% and 36%, respectively, higher mortality risks than those of patients treated in units with more than 45 stations. Conclusion: A small facility size was associated with an increased mortality risk in HD patients, and the threshold size was higher in freestanding units.