Winnipeg, Canada operates a 16-bed subacute unit, the Crisis Stabilization Unit (CSU), for voluntary patients in crisis not requiring hospital admission. The virtual CSU (vCSU) launched in March 2020 as an adjunct to the in-person CSU during the COVID-19 pandemic, providing the same resources virtually, allowing patients to remain at home. Program data were collected for vCSU admissions between April 1, 2020 and April 7, 2021 (n = 266) to examine patient characteristics and discharge outcomes. Data were retrieved from the electronic patient record (EPR) for both in-person and vCSU admissions during the same period for comparison (n = 712). vCSU admissions (n = 191) were summarized by patient demographics, clinical factors/outcomes, and compared on the same measures to in-person CSU admissions (n = 521) using binary logistic regression. 30.1% of patients admitted to the vCSU received initial mental health assessment virtually (phone/videoconference), therefore receiving all care at home. Clinical symptoms at assessment included depression/anxiety (39.0%), psychosis/mania (2.7%), suicidal behaviour/self-harm (27.4%), psychosocial event/stressor (19.8%). Average stay was 4.9 days. Compared to the in-person CSU, vCSU referrals were associated with the absence of psychosis [odds ratio (OR).40, 95% confidence interval (CI).18-0.89] and no prior 1-year contact with referral site (OR.43, 95% CI.28-0.64). Those living farther away from the referral site were more likely to receive a vCSU referral. The vCSU model is feasible for a diverse group of patients experiencing mental health crises. Future work is needed to better determine who the model is right for and examine longer term outcomes.