Background: Exploring the association between the electrocardio-gram (ECG) and cardiac magnetic resonance (CMR)-derived features may enhance our understanding of cardiovascular physiology and enable more detailed evaluation of the heart's structural characteristics. We aimed to identify clusters of individuals without diagnosed cardiovascu-lar disease (CVD) based on their ECG phenotypes in an unsupervised manner and evaluate their cardiac anatomical differences through CMR. Methods: Spatial and single-lead ECG biomarkers (N=242) were cal-culated from 10-second 12-lead ECGs from 51,974 UK Biobank individ-uals without diagnosed CVD. ECG phenotypes, adjusted for age, sex, and body mass index, were clustered using an unsupervised k-means model. Statistical analyses were conducted to assess ECG, demographical and CMR differ-ences across clusters, and multivariable linear regressions evaluated the electro-anatomical alignment within each cluster. Results: Three distinct ECG-based clusters were identified (N1=19,470, N2=22,256, N3=8,997), with significant differences in ECG morphology and CMR-derived anatomical features. The most discrimina-tive ECG features involved ventricular repolarization in precordial leads (QRS-, T- and ST-segment amplitudes and T-wave morphological varia-bility index). Cluster 3 showed the strongest electro-anatomical align-ment, with right ventricular end-systolic and end-diastolic volumes con-tributing importantly (or driving much) of the ECG variation across clus-ters. Conclusions: Our findings show that ECG phenotyping through unsu-pervised clustering can reveal anatomical cardiac differences, offering new perspectives on early cardiovascular variation in the general popu-lation. Future research should assess the longitudinal and prognostic implications of these phenotypic clusters.