This study uses satelite data to estimate gross regional domestic product (GRDP) in North Korea and examines the spatial distribution, level, and structure of regional disparities. For 210 cities, counties, and districts from 1992 to 2020, value added outside agriculture, forestry, and fisheries was allocated using nighttime light data, while agricultural, forestry, and fisheries value added was estimated by integrating land-cover data with a digital elevation model (DEM). The applicability of the estimation method was indirectly assessed using official regional statistics for Jilin Province, China. Regional disparities were analyzed using a three-stage nested decomposition combining Dagum’s Gini decomposition with Akita’s nested structure. The results reveal a western inland axis centered on Pyongyang, an eastern coastal axis of heavy-industrial cities, and border nodal areas characterized by external linkages. In 2020, North Korea’s GRDP Gini coefficient was 0.51, compared with 0.32 in South Korea, and disparities between the city and county groups accounted for approximately 69% of total regional inequality. These spatial patterns may be associated with institutional characteristics identified in previous studies, including concentration in Pyongyang, heavy-industry-oriented industrial location, and the local self-reliant economic system. This study contributes by constructing long-term city- and county-level economic estimates and quantifying the contribution of different spatial hierarchies to regional disparities.