This dataset contains annual plastic greenhouse (PG) classification maps for Northeast China from 2019 to 2025, provincial boundary data for the study area, and reference sample data used for accuracy assessment. The PG maps were generated on the Google Earth Engine platform using Sentinel-1 SAR GRD, Sentinel-2 MSI Harmonized Surface Reflectance, and ancillary spatial constraints. The annual PG maps are provided as GeoTIFF files, where pixel value 1 indicates plastic greenhouse and pixel value 0 indicates non-plastic greenhouse. The file PG-2019-2025.zip contains the annual PG classification maps for 2019–2025. The file NEC.zip contains the provincial boundary shapefile for Northeast China used to define the study area. The file All_sample.csv contains the 2022 reference samples, including longitude, latitude, and land-cover class labels. The sample class codes are: 1 = plastic greenhouse, 2 = cropland, 3 = forest, 4 = grassland, 5 = built-up land, and 6 = water.