Objective: The coronavirus disease 2019 (COVID-19) has placed huge strains on medical systems. Therefore, it is essential to determine the predictors of the long hospital stay. We sought to investigate whether alterations in left ventricular (LV) geometry in COVID-19 patients are associated with the length of stay (LoS) and a long hospital stay. Materials and Methods: 108 consecutive hospitalized COVID-19 patients were incorporated in the study and 89 patients remained for statistical analysis. All participants underwent standard two-dimensional (2D) and Doppler echocardiographic examinations. Patients were classified according to LV geometry characteristics namely normal geometry (NG), concentric remodeling, concentric hypertrophy and eccentric hypertrophy. Results: Multiple binary logistic regression model adjusted for clinical and laboratory variables yielded significant and independent association of LV mass index (LVMI) (OR: 1.12, 95% CI: 1.06–1.19, p<0.001), 10 g/m2 increase in LVMI (OR: 3.63, 95% CI: 2.00–6.59, p<0.001), LV geometry patterns (OR: 2.92, 95% CI: 1.46–5.84, p=0.002), and altered geometric patterns compared to NG (OR: 3.97, 95% CI: 1.08–14.5, p=0.037) with long hospital stay. Correlation analysis of LVMI and LoS demonstrated significant and moderate correlation (rho=0.58, p<0.001). Conclusion: LVMI and LV geometric patterns independently predict long hospital stays in COVID-19 patients. The significant correlation between LoS and LVMI underlies the significance of LV geometry in this infection.