Acute lung injury is associated with dysfunctional immune response to SARS-CoV-2. This leads to CRS, which require immunomodulatory treatments aiming to limit the excessive production of cytokines. The literature so far indicates the effectiveness of tocilizumab in patients with COVID-19-associated pneumonia, but there is no clear evidence of its effectiveness in patients with at least 50% lung involvement; therefore, we aimed to bridge this gap in knowledge. Longitudinal data for 4287 patients with confirmed COVID-19 infection were collected between 1st March 2020 and 16th of January 2022. In total, 182 cases with lung involvement >50% and biochemical indicators of cytokine release storm (Il-6 >100 pg/mL) were selected and analyzed using non-parametric statistics and multivariate Cox models. Among the 182 included patients, 100 (55%) were treated with TCZ, while 82 (45%) did not receive TCZ. The groups were balanced regarding demographics, lung involvement and biochemical markers. Overall mortality in the group was 63.1%. Mortality in the TCZ group was 58.0% compared to 69.5% (n = 57) in the non-TCZ group (p = 0.023). In multivariate Cox proportional hazards models, intravenous administration of tocilizumab was associated with lower probability of ICU admission (HR: 0333 (CI: 0.159-0.700, p = 0.004)) and lower mortality (HR: 0.57306 (CI: 0.354-0.927, p = 0.023)). Tocilizumab is effective as a treatment in the most severely ill patients, in whom the level of lung involvement by the inflammatory process can exceed 50% with coexisting biochemical indices of cytokine storm (Il-6 > 100 pg/mL).