Abstract: There is a strong correlation between light data and human social and economic activities. Therefore, building a GDP prediction model based on light data and GDP spatialization has important research value. Taking the Pearl River Delta region as an example, the night light data of Luojia No.1 were preprocessed, including mosaic registration, intra-year stability correction, mask denoising, and radiation correction, etc., which aims to obtain the 2018 Pearl River Delta LJ1-01 night light data set; Based on the night light data, the spatial distribution characteristics were analyzed, and the total light radiation value of each administrative area in the Pearl River Delta region was extracted for hot spot analysis; In addition, based on the night light data and GDP statistics, four lighting indices (TNL, I, S, CNLI) for the districts and counties of the administrative area were extracted, and the linear, logarithmic, power index, and exponential models with GDP were constructed. The best lighting index and the best regression model were determined through the determination coefficients; Based on the optimal lighting indicators and the best regression model, error correction was performed to obtain a spatial density map of GDP for the Pearl River Delta in 2018. Research results show that the small-scale grid format data obtained based on the GDP spatialization of the LJ1-01 data can reflect the region's spatial differences and spatial accuracy in a more detailed way than NPP-VIIRS data.