Background and Purpose: The water-swallowing test (WST) is frequently used in clinical practice as a functional assessment to detect aspiration and prevent pneumonia. It is a standardized test used all over the world, but the amount of water given varies depending on the examiner. Furthermore, there are very few reports on the simultaneous performance of the WST and videofluorography (VF). This study compared the amount of swallowed water to investigate the reliability of WST to exclude aspiration following acute stroke. Methods: We assessed 111 stroke patients (65 men and 46 women) with suspected dysphagia/difficulty in swallowing and performed VF upon obtaining consent from the patients and their families. Patients were aged between 20 and 98 years (65.6 ± 13.4 years); 64 had cerebral infarction, 26 cerebral hemorrhage, 13 subarachnoid hemorrhage, and 8 had other cerebrovascular disease. The time from stroke onset to VF was 16.6 ± 10.3 days (range, 2-55). WSTs using 5, 10, 30, and 60 ml and the modified WST (MWST) were performed during VF. Results: We found that the number of instances of choking, cough, wet voice, and aspiration increased with higher amounts of water. The sensitivity and specificity of WST for aspiration ranged from 34.8 to 55.7% and from 78.9 to 93.2%, respectively. The MWST, which used only 3 ml of water, yielded a sensitivity of 55.3% and a specificity of 80.8% for aspiration. There was a positive correlation between the time for one swallow and age, but there was no difference between genders. There was also no connection between clinical findings during WST or the presence of aspiration with the number of swallows, swallowing speed, or time for one swallow. Conclusions: WSTs are not as powerful as VF as a screening instrument in acute stroke. WSTs with more water detected aspiration with greater sensitivity, but there is no justification for overconfidence when investigating aspiration. We recommend using WST as well as VF to investigate swallowing in stroke patients.