Objective To investigate the correlation between gait speed and mortality in very old inpatients. Methods A total of 510 very old inpatients (≥80 years) at our department from June 2015 to December 2017 were enrolled in this study. Their basic data, comorbidities and polypharmacy were recorded at admission. Their activities of daily living was assessed by Katz index of the Katz in activities of daily living (Katz-ADL), cognitive function by mini-mental state examination (MMSE), nutrition level by micronutrient assessment method short-form (MNA-SF), and depression status by self-rating depression scale (SDS). The gait speed was measured with 6-meter walking test. After discharge, all-cause mortality was followed up for more than 2 years and recorded. SPSS statistics 21.0 was used for data analysis. Spearman correlation coefficient was used to analyze the correlation between gait speed and variables. Kaplan-Meier survival curve and Cox regression analysis were used to analyze the correlation between gait speed and mortality.Results For the 499 inpatients, they were at a mean age of (86.3±4.1) years, and 59.7% of them were male. During a median follow-up of 3.3 years (2.7,3.9), 118 participants (23.6%) died. Spearman correlation analysis revealed that gait speed was negatively correlated with age, Charlson comorbidity index (CCI) and polypharmacy (r=-0.352, -0.196, -0.124; P<0.05), and positively correlated with Katz-ADL, MMSE and MNA-SF scores (r=0.430,0.291,0.273; P<0.05). After the patients were divided into tertile groups according to gait speed, Kaplan-Meier survival analysis showed that mortality was increased with the decrease of gait speed (P<0.001). Cox regression analysis with adjustment for potential confounding variables indicated that compared with the fastest group, the other 2 groups have a higher risk of mortality (middle speed group:HR=2.365,95%CI 1.231～4.541, P=0.010; lowest speed group:HR=2.363,95%CI 1.215～4.591, P=0.011). Conclusion In elderly inpatients, low gait speed is associated with the risk of all-cause mortality. Gait speed could be a simple and effective prognostic indicator for mortality.