To clarify the correlation between malnutrition, muscle mass and oral status, and swallowing function recovery in stroke patients receiving enteral nutrition.Patients with stroke and dysphagia receiving any amount of enteral nutrition in rehabilitation wards from 2012 to 2016 were eligible for inclusion in this retrospective study. On admission, body composition by bioimpedance analysis, malnutrition confirmed by the European Society for Clinical Nutrition and Metabolism criteria, oral status, functional independence measure and demographic data were collected. Characteristics were compared between "oral intake alone" and "artificial nutrition" groups based on the discharge status. Kaplan-Meier methods and the Cox proportional hazards model were used to determine explanatory factors for the probability of full oral intake.Among 174 patients, 113 were analysed (55 women; median age, 77 years). Overall, 61% and 39% were classified as "oral intake alone" and "artificial nutrition," respectively. Days from onset to admission to rehabilitation wards and motor Functional Independence Measure were higher in the "oral intake alone" group. Kaplan-Meier analysis demonstrated that patients with lower muscle mass exhibited lower probability of full oral intake (P = .009). The Cox proportional hazards model suggested that lower muscle mass (hazard ratio, 0.493; 95% CI, 0.286-0.850) and poor oral hygiene (hazard ratio, 0.573; 95% CI, 0.333-0.987) were independently correlated with "oral intake alone" status. Malnutrition and other oral status are not related to achieving full oral intake.Skeletal muscle mass and oral hygiene are independently correlated with full oral intake among stroke patients receiving enteral nutrition during the rehabilitation phase.