Aims: The aim of this study was to investigate the association between Type 2 diabetes and community-acquired infections. Methods: We recruited 68 patients (mean +/- S.D. age=64.9 +/- 9.1 years; 57.4% males) from a community-based cohort and their nondiabetic partners (age=63.8 +/- 9.9 years; mean relationship duration=37.3 +/- 11.5 years) to a prospective observational matched-pair cohort study. Participants were assessed for infection risk factors and were required to complete an infection diary over the following year. Results: Sixty patient-partner pairs returned completed diaries. A greater proportion of patients than partners had received influenza vaccination in the past year (73.3% vs. 60.0%; P=.057), but pneumococcal vaccination coverage in the previous 5 years was similar (11.7% vs. 10.0%; P= 1.0). The proportions of patients and partners reporting no infections were similar (51.7% vs. 58.3%; P=.52), but the diabetic patients were more likely to have experienced more than one infection (33.3% vs. 18.3%: P=.022). Baseline glycosylated hemoglobin levels in patients with and without at least one infection were similar, and there was no association between multiple infections and glycosylated hemoglobin (P>.2 in each case). After adjusting for age and gender, the probability of at least one infection during follow-up was not associated with diabetes (P=.33), but the number of infections was positively associated with diabetes (P=.005). Conclusion/interpretation: Although similar proportions of Type 2 diabetic patients and their nondiabetic partners remained infection-free over a 12-month period, the diabetic patients who developed infections had a greater number of episodes. Glycemic control did not predict the risk or frequency of community-acquired infections in the diabetic group. (C) 2005 Elsevier Inc. All rights reserved.