Smoking can play a key role in SARS-CoV-2 infection and in the course of the disease. Previous studies have conflicting or inconclusive results on the prevalence of smoking and the severity of the coronavirus disease (COVID-19). Observational, multicenter, retrospective cohort study of 14,260 patients admitted for COVID-19 in Spanish hospitals between February and September 2020. Their clinical characteristics were recorded and the patients were classified into a smoking group (active or former smokers) or a non-smoking group (never smokers). The patients were followed up to one month after discharge. Differences between groups were analyzed. A multivariate logistic regression and Kapplan Meier curves analyzed the relationship between smoking and in-hospital mortality. The median age was 68.6 (55.8-79.1) years, with 57.7% of males. Smoking patients were older (69.9 59.6-78.0 years), more frequently male (80.3%) and with higher Charlson index (4 2-6) than non-smoking patients. Smoking patients presented a worse evolution, with a higher rate of admission to the intensive care unit (ICU) (10.4 vs 8.1%), higher in-hospital mortality (22.5 vs. 16.4%) and readmission at one month (5.8 vs. 4.0%) than in non-smoking patients. After multivariate analysis, smoking remained associated with these events. Active or past smoking is an independent predictor of poor prognosis in patients with COVID-19. It is associated with higher ICU admissions and in-hospital mortality.