OBJECTIVE Percutaneous endoscopic gastrostomy feeding provides enteral nutrition to patients with neurological dysphagia. Thirty-day mortality rates of 4-26% have been reported, with pneumonia being the common cause post-percutaneous endoscopic gastrostomy insertion. METHODS This retrospective analysis of percutaneous endoscopic gastrostomy tube insertions in Malta (January 2008 - June 2010) compares the incidence of pneumonia in patients fed through a nasogastric tube versus in those fed via a percutaneous endoscopic gastrostomy tube. We analyzed the indications, poor prognostic factors and mortality for percutaneous endoscopic gastrostomy insertion. RESULTS Ninety-seven patients underwent percutaneous endoscopic gastrostomy insertion. Fifty-four patients received nasogastric feeds before percutaneous endoscopic gastrostomy feeds. Patients on nasogastric feeds developed 32 episodes of pneumonia over a total of 7884 days of feeds (1 every 246 days). Patients with percutaneous endoscopic gastrostomy feeds after a period of nasogastric feeds developed 48 pneumonia episodes over 36,238 days (1 every 755 days). Patients with percutaneous endoscopic gastrostomy feeds without previous nasogastric feeds developed 28 pneumonia episodes over 23,983 days (1 every 856 days), and this was statistically significant (χ 2 test p value <0.005). Forty-seven patients had died at the time of data collection, with 29 patients dying from pneumonia. One-week mortality was 3%, 30-day mortality was 8% and 1-year mortality was 39%. All patients dying within the first week and 50% of those dying within 30 days of the procedure died following pneumonia. CONCLUSIONS There was a statistically significant decrease in the number of pneumonia episodes among patients receiving percutaneous endoscopic gastrostomy feeds versus nasogastric feeds. However, pneumonia is still the major cause of death among percutaneous endoscopic gastrostomy patients.