This study investigates the influence of the North Atlantic Oscillation (NAO) phases on autumn and winter temperatures in East Azerbaijan Province, using observational data from eight selected meteorological stations over a 20-year period (2004–2023). Monthly temperature data and NAO indices were acquired from the NOAA database. Correlation coefficients and statistical significance were computed using the R environment, while the spatial patterns of these teleconnections were mapped using Geographic Information Systems (GIS). The results indicated no statistically significant relationship between the NAO index and winter temperatures in the region. Conversely, a significant inverse correlation was observed during the autumn season. This relationship was statistically significant at the Kaleybar, Ahar, and Marand stations, while the Sarab and Miane stations exhibited similar, albeit near-threshold, trends. The peak correlation was identified at the Kaleybar station, with an inverse coefficient of approximately -0.6 (p ≤ 0.01). These findings highlight the heterogeneous and spatially dependent nature of autumn temperature responses to the NAO within the study area. Specifically, a positive NAO phase is associated with lower autumn temperatures, whereas a negative phase corresponds to warming. The prevalence of significant correlations in the eastern part of the province suggests a heightened sensitivity of local agricultural systems in these areas to NAO-driven climate variability.