Objective: To evaluate the effect of influenza vaccination on the reduction of the risk of outpatient visits for upper respiratory infection (URI) among the elderly in Taiwan. Methods: The data for this observational study, consisting of 1729 people aged 65 years or older, were drawn from Taiwan's 2001 National Health Interview Survey. This survey data was then linked with National Health Insurance claim data for December 2001 to November 2002. Survival analysis of Cox proportional hazards model was performed to examine the risk of URI outpatient visits in elderly people vaccinated with the influenza vaccine and those not vaccinated during a year-long study period since the influenza season began. To adjust for potential self-selection bias, we used propensity score method to categorize individuals into two groups, based on the predicted probability of being vaccinated from a logistic regression of spatial random effect. Propensity score group 1 (PSG 1) were those with a predicted probability of being vaccinated lower than 0.5, and PSG 2 were those with a predicted probability of being vaccinated of 0.5 or higher. Results: The overall vaccination rate was 50%. Logistic regression showed the probability of being vaccinated was related to the number of outpatient visits for URI before the influenza season began (odds ratio (OR) 1.07; 95% confidence interval (CI) 1.04-1.10). Our first survival analysis showed that being vaccinated significantly reduced the risk of URI outpatient visits in PSG 2 during the 1-year study period (hazard ratio 0.89; 95% CI 0.81-0.97). Separate survival analysis showed that being vaccinated reduced the risk of URI outpatient visits for both PSG groups during the first 3 months of the study period. Conclusion: Being vaccinated could reduce the risk of outpatient visits for URI among the elderly during the influenza season.