Abstract Aims The recent 4S-AF (scheme proposed by the 2020 ESC AF guidelines to address stroke risk, symptom severity, severity of AF burden and substrate of AF to provide a structured phenotyping of AF patients in clinical practice to guide therapy and assess prognosis) scheme has been proposed as a structured scheme to characterize patients with atrial fibrillation (AF). We aimed to assess whether the 4S-AF scheme predicts AF progression in patients with self-terminating AF. Methods and results We analysed 341 patients with self-terminating AF included in the well-phenotyped Reappraisal of Atrial Fibrillation: Interaction between HyperCoagulability, Electrical remodelling, and Vascular Destabilization in the Progression of AF (RACE V) study. Patients had continuous monitoring with implantable loop recorders or pacemakers. AF progression was defined as progression to persistent or permanent AF or progression of self-terminating AF with >3% burden increase. Progression of AF was observed in 42 patients (12.3%, 5.9% per year). Patients were given a score based on the components of the 4S-AF scheme. Mean age was 65 [interquartile range (IQR) 58–71] years, 149 (44%) were women, 103 (49%) had heart failure, 276 (81%) had hypertension, and 38 (11%) had coronary artery disease. Median CHA2DS2-VASc (the CHA2DS2–VASc score assesses thromboembolic risk. C, congestive heart failure/left ventricular dysfunction; H, hypertension; A2, age ≥ 75 years; D, diabetes mellitus; S2, stroke/transient ischaemic attack/systemic embolism; V, vascular disease; A, age 65–74 years; Sc, sex category (female sex)) score was 2 (IQR 2–3), and median follow-up was 2.1 (1.5–2.6) years. The average score of the 4S-AF scheme was 4.6 ± 1.4. The score points from the 4S-AF scheme did not predict the risk of AF progression [odds ratio (OR) 1.1 95% CI 0.88–1.41, C-statistic 0.53]. However, excluding the symptoms domain, resulting in the 3S-AF (4S-AF scheme without the domain symptom severity, only including stroke risk, severity of AF burden and substrate of AF) scheme, predicted the risk of progression (OR 1.59 95% CI 1.15–2.27, C-statistic 0.62) even after adjusting for sex and age. Conclusions In self-terminating AF patients, the 4S-AF scheme does not predict AF progression. The 3S-AF scheme, excluding the symptom domain, may be a more appropriate score to predict AF progression. Trial registration numbers Clinicaltrials.gov NCT02726698 for RACE V