At the beginning of the COVID-19 pandemic, there was high mortality and a lack of effective treatment for critically ill patients. Build on the experience in argentine hemorrhagic fever with convalescent plasma, we incorporated 90 patients into a multicenter study, and 87 were evaluable. We collected 397 donations from 278 convalescent donors. Patients received plasma with an IgG concentration of 0.7-0.8 (measured by Abbott chemiluminescence) for every 10 kg of body weight. Survival during the first 28 days was the primary objective. 77% were male, age 54 (+/-15.6 y/o (range 27-85); body mass index 29.7 +/-; 4,4; hypertension 39% and diabetes 20%; 19.5% had an immunosuppression condition; 23% were healthcare workers. Plasma was administered to 55 patients (63%) on spontaneous breathing with oxygen supplementation (mainly oxygen mask with reservoir bag in 80%), and 32 patients (37%) were infused on mechanical ventilation. The 28-day survival rate was 80%, with 91% in patients infused on spontaneous breathing and 63% in those infused on mechanical ventilation (p = 0.0002). There was a significant improvement in the WHO pneumonia clinical scale at 7 and 14 days, and in PaO2 / FiO2, ferritin and LDH, in the week post-infusion. We observed an episode of circulatory volume overload and a febrile reaction, both mild. Convalescent plasma infusions are feasible, safe, and potentially effective, especially before requiring mechanical ventilation, and are an attractive clinical option for treating severe forms of COVID-19 until other effective therapies become available.