Objective: Determine whether the tomographic characteristics of patients with COVID-19 pneumonia at hospital admission and the initial tomographic severity score (TSS) as well as some laboratory tests or clinical characteristics predict mortality and length of hospital stay.Material and methods: Analytical, retrospective study that included 203 patients with a clinical diagnosis of SARSCoV2 Virus Infection, performed by reverse transcriptase polymerase chain reaction (RT-PCR), reactive serological test (IgM/IgG) and/or spiral computed tomography (CT) chest without contrast, admitted in the period from 6 April to 27 June 2020. Two radiologists (blind evaluators) described the tomographic findings. 15 patients with normal CT were excluded, 188 patients with clinical and laboratory parameters were admitted to the statistical analysis of TSS and mortality; data taken from epidemiological records and clinical histories, divided into two groups: recovered and deceased. Data analysis used the statistical package Stata version 14.2; hospital stay was analyzed by Kaplan – Meier curves and mortality prediction was made by Cox regression with 95% CI and a p value <0.05 as statistically significant.Results: Hypertension was the most frequent associated disease, the most common clinical presentation included cough, malaise, fever and dyspnea , there were no significant differences between studied groups   (Recovered  vs. death patients), 15 cases had normal CT, for this reason, they were excluded from the TSS score and mortality analysis. Mean age of the recovered patients was 53.6 ± 16.4 years and of the deceased 75.9 ± 13.9 years (p <0.0001). A moderate and high TSS (≥ 8) resulted in deaths (p <0.05), as well as a higher degree of lymphopenia and a history of asthma in the deceased (p <0.05). The variable arterial hypertension predicts the increase in days of hospital stay (p <0.05). The ground glass pattern was the most frequent, followed by the consolidation and distortion of the architecture; however, they were not associated with a type of outcome. The pattern of pleural effusion and bronchial dilation showed a significant difference with respect to the outcome, however, but it could be due to the infrequency of presentation of these patterns. Conclusions: TSS is useful in the initial and comprehensive diagnostic evaluation of COVID-19 pneumonia, in conjunction with markers such as lymphopenia and elevated CRP that can predict a poor short-term outcome. A high TSS score is a predictor of mortality. Hypertension predicts the increase in days of hospital stay.