A total of 14,071 earthquakes with magnitude ≥ 6 occurring between 1900 and 2024 were analysed, with the aim of exploring whether their occurrence exhibits patterns associated with angular separations between astronomical positions. To this end, the full range of angular separations was systematically examined using an ephemeris-based control model and independently validated through Monte Carlo simulations. The analysis identifies a limited number of angular regions in which a relative concentration of events is observed, with statistically significant deviations from those expected under a temporally random distribution. These regions remain stable when varying the start degree and width of the angular intervals and when applying different statistical controls. Although the observed effects are small and do not support predictive or causal inferences, they reveal reproducible regularities in the distribution of angular separations. The study provides a quantitative framework for investigating possible systematic variations in astronomical angular distances associated with the occurrence of earthquakes, and establishes a methodological basis for future research with higher spatial and geodynamic resolution.