Background: Research suggests NT-proBNP (N-terminal pro-B-type natriuretic peptide) to be a strong predictor of incident atrial fibrillation (AF) and stroke. However, its utility in AF screening remains unknown. The aim of this study was to investigate NT-proBNP as a potential marker for screening efficacy with respect to AF yield and stroke prevention. Methods: In the LOOP Study (Atrial Fibrillation Detected by Continuous ECG Monitoring Using Implantable Loop Recorder to Prevent Stroke in High-Risk Individuals), 6004 AF-naïve individuals at least 70 years old and with additional stroke risk factors were randomized 1:3 to either screening with an implantable loop recorder (ILR) and initiation of anticoagulation upon detection of AF episodes lasting ≥6 minutes or usual care (control). This post hoc analysis included study participants with available NT-proBNP measurement at baseline. Results: A total of 5819 participants (96.9% of the trial population) were included. The mean age was 74.7 years (SD, 4.1 years) and 47.5% were female. The median NT-proBNP level was 15 pmol/L (interquartile range, 9–28 pmol/L) corresponding to 125 pg/mL (interquartile range, 76–233 pg/mL). NT-proBNP above median was associated with an increased risk of AF diagnosis both in the ILR group (hazard ratio, 1.84 [95% CI, 1.51–2.25]) and the control group (hazard ratio, 2.79 [95% CI, 2.30–3.40]). Participants with NT-proBNP above the median were also at higher risk of clinical events compared with those having lower levels (hazard ratio, 1.21 [95% CI, 0.96–1.54] for stroke or systemic embolism [SE], 1.60 [95% CI, 1.32–1.95] for stroke/SE/cardiovascular death, and 1.91 [95% CI, 1.61–2.26] for all-cause death). Compared with usual care, ILR screening was associated with significant reductions in stroke/SE and stroke/SE/cardiovascular death among participants with NT-proBNP above median (hazard ratio, 0.60 [95% CI, 0.40–0.90] and 0.70 [95% CI, 0.53–0.94], respectively) but not among those with lower levels ( P _interaction =0.029 for stroke/SE and 0.045 for stroke/SE/cardiovascular death). No risk reduction in all-cause death was observed in either NT-proBNP subgroup for ILR versus control ( P _interaction =0.68). Analyzing NT-proBNP as a continuous variable yielded similar findings. Conclusions: In an older population with additional stroke risk factors, ILR screening for AF was associated with a significant reduction in stroke risk among individuals with higher NT-proBNP levels but not among those with lower levels. These findings should be considered hypothesis generating and warrant further study before clinical implementation. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT02036450.