Abstract Background Primary prevention of atrial fibrillation (AF) is important given its rising incidence, associated disease burden, and cost (1-3). The effect of pharmacological therapies on incidence of new-onset AF is uncertain, (1) leading to an absence of clear guideline recommendations for prevention of AF (3). Purpose To perform a meta-analysis of the effect of pharmacological therapies on new-onset AF as reported in published randomised clinical trials (RCTs), and investigate for changes by individual agents and clinically relevant subgroups. Methods We performed a systematic review of RCTs that evaluated the relationship between a pharmacological therapy and the development of new-onset AF in a non-hospitalised patient cohort, using Medline, Embase and Cochrane-controlled Trials Register from inception through 19 July 2023. Results of the AF outcome are expressed as relative ratio (RR) with 95% confidence interval (CI) between the intervention and control group for each study. A fixed-effect model was used to combine results from studies when I2<50% and random-effect model when I2>50%. Risk of bias was assessed using the Cochrane RoB2 tool. Meta-regression was performed by mean age of study participants, percentage of men, follow-up duration, indication (e.g. heart failure) and each agent. Results From 11872 unique records we included 101 RCTs with 794775 patients. Overall, risk of bias was considerable (48% high, 35% medium); driven by outcome measurement (67% high/medium) as AF was not the primary outcome of studies. Of pharmacological therapies, patients treated with angiotensin converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) demonstrated a lower risk of AF compared with controls (RR 0.87, 95% CI 0.78-0.98), and patients treated with omega-3 fatty acids demonstrated a higher risk of AF (RR 1.26, 95% CI 1.11-1.42) (Figure 1). Analysis by indication demonstrated that the reduction in AF risk with treatment with ACEIs/ARBs was only for patients with heart failure with reduced ejection fraction (HFrEF) (RR 0.57, 95% CI 0.34-0.94). Amongst individual ACEIs/ARBs, only treatment with valsartan (RR 0.79, 95% CI 0.66-0.95) and olmesartan (RR 0.87, 95% CI 0.45-0.99) led to a lower risk of AF. On meta-regression each year of increased age increased AF risk when treated with ACE-I/ARB therapy (RR 1.08, 95% CI 1.04-1.13). Whilst no reduction in AF was observed for other pharmacological classes overall, treatment with Semaglutide demonstrated a 40% lower risk of AF compared with controls (RR 0.60, 95% CI 0.45-0.80) (Figure 2). Conclusions ACEIs and ARBs appear to be effective in the primary prevention of AF, but this benefit may be limited to patients with HFrEF, especially younger patients. Effect on AF occurrence varies within pharmacological classes, and Semaglutide specifically may be an effective in preventing new-onset AF. Appropriately powered RCTs with new-onset AF as the primary outcome are required.Figure 1:Forest plot by pharmacotherapyFigure 2:Forest plot by GLP-1RA agent