Abstract Introduction Bacterial colonization of generator pockets in implantable cardioverter defibrillator (ICD) patients plays a key role in the development of device infection after revision surgery. Pocket swabs taken at revision may be useful for predicting device infection, but their prognostic value is not precisely known. Aim of the study was to evaluate the predictive value of positive pocket swabs for cardiac device infection in ICD patients undergoing revision surgery. Methods A total of 646 consecutive ICD patients from a prospective single-centre ICD-registry who underwent elective device replacement, system upgrade or lead revision between 2014 and 6/2023 were analyzed. Independent predictors for device infection were evaluated by a Cox regression analysis adjusting for age, gender, EF, insulin dependent diabetes, anticoagulation, multiple leads, revision surgery and the PADIT score. Results Positive pocket swabs were obtained in 12% of patients. The baseline characteristics of ICD patients with positive versus negative pocket swabs were similar (Table 1). With an average follow-up of 3 years, the infection rate in patients with positive pocket swabs was around 5 times higher than in patients with negative swabs (19% versus 4%, p < 0.001). In a Cox regression analysis, a positive pocket swab was the strongest independent predictor for device infection during follow-up (hazard ratio 4.26, 95% confidence interval 2.15 – 8.45, Table 2). Conclusion A positive swab from ICD-pockets in patients undergoing generator replacement or lead revision is an independent predictor of device infection during follow-up. These results support the routine collection of pocket swabs from patients undergoing revision surgery.Table 1Table 2