Introduction: Pneumonia contributes significantly to the morbidity and mortality in stroke patients, especially those fed via nasogastric tubes. Methods: This project was conducted in two steps; 1. A randomised controlled trial: The efficacy of prokinetic agent metoclopramide was tested in a double-blind randomised controlled trial. Acute stroke patients with no pneumonia needing nasogastric feeds were randomized to 10 mg metoclopramide or placebo three times daily via the nasogastric tube for 21 days or until feeds discontinued. Participants were examined daily for clinical evidence of pneumonia. 2. A secondary analysis of data collected for the diagnosis of pneumonia. This was performed to identify early diagnostic markers of post-stroke pneumonia. Results: 1) For the MAPS study 60 patients (mean age 78 years, mean NIHSS 19) were randomized, 30 in each group. Pneumonia was diagnosed in 26/30 and 08/30 in placebo and treatment group respectively (p < 0.001). 2) Signs and symptoms of 47 radiologically confirmed pneumonia revealed that hypoxia, tachypnoea and inspiratory crackles were the most consistent signs in early pneumonia and pyrexia, cough, and purulent sputum were less commonly observed. CRP of 40 mg/l had the best predictive value for early diagnosis of post-stroke pneumonia. Conclusion: Treatment with metoclopramide treatment was associated with a significant reduction of aspiration and pneumonia. However, the incidence of pneumonia in the control gourp was very high and may have led to a false positive result. The findings of this study should therefore be confirmed in a larger study. A raised CRP of 40 mg/l or above was a senditive and specific diagnostic marker for post-stoke pneumonia.