Background How and where to initiate dialysis are policy challenges with enormous economic and health consequences. Initiating with home hemodialysis (HD) or peritoneal dialysis (PD) may reduce costs and improve outcomes but evidence is conflicting. Methods We conducted a population-based study in patients aged ≥ 18 years who initiated chronic dialysis in the province of Ontario, Canada from 2006 to 2014 ( N = 12,691) using linked administrative data. Patients were grouped by initial modality: facility HD, facility short daily or slow nocturnal (SD/SN) HD, PD, home HD. We estimated publicly-paid healthcare costs (2015 Canadian dollars; 1 = 0.947 US dollar) and survival, from dialysis initiation to March 2015. Results By 5 years after dialysis initiation, mean 30-day costs (as-treated) for patients receiving PD and home HD were 50% and 64% lower, respectively, than for facility HD patients ($11,011). Approximately 50% of costs were unrelated to dialysis, reflecting high comorbidity in these patients. With covariate adjustment, mean 5-year cumulative costs were similar for initiators of home HD and PD ($304,178 and $349,338) and higher for facility HD initiators ($410,981). The highest 5-year unadjusted survival was for home HD patients (80%), followed by PD (52%), SD/SN HD (50%), and facility HD (42%). Conclusions This study in a large cohort over 9 years provides new population-based evidence suggesting that initiating dialysis at home is cost-effective, with lower costs and better survival, than starting with facility HD. Survival differences persisted after adjustment for baseline characteristics but we could not adjust for functional status or severity of comorbidities.