Background: To investigate whether fluoxetine improves poststroke dysphagia and to detect the potential relationship between serum brain-derived neurotrophic factor (BDNF) levels and fluoxetine effects. Methods: In this retrospective study, 159 stroke patients who met our study criteria were included. In total, 110 patients were placed in the control group, and 49 patients were placed in the fluoxetine group. Demographic and clinical characteristics of the patients were collected for the baseline assessment. Functional independence measure scores and American speech-language-hearing association/functional communication measures scores for swallowing were collected to evaluate the patients' swallowing function. Patients' serums were collected at weeks 1 and 3 after admission, and serum BDNF levels were measured by enzyme-linked immunosorbent assay. T test, chi-squared test, and general linear model analysis were performed to determine the differences between the two groups. Results: A significantly higher improvement of swallowing function was observed in the fluoxetine group compared with that of the control group (P = .023). In addition, a general linear model analysis showed that the treatment of fluoxetine has a statistically significant effect on swallowing improvement after adjustment of swallowing score on admission, stroke types, and interval between the onset of stroke and admission (P = .022, R-2 = .46, adjusted R-2 = .446). There is no significant difference in the change of serum BDNF levels in the two groups (P = .269). Conclusions: This study suggests that treatment with fluoxetine in stroke patients with dysphagia may improve swallowing function. A placebo-controlled, randomized clinical trial is warranted to confirm this finding.