A county-level spatiotemporal dataset of land transfer disputes in China

This dataset contains document-level records of land transfer disputes in China derived from China Judgments Online. The dataset was constructed through keyword-based retrieval, topic screening, DeepSeek-R1-assisted field extraction, case-number linkage, case clustering, variable standardization and manual correction. The main dataset contains 305,256 judicial documents corresponding to 223,280 unique land transfer dispute cases. It covers the period 2002–2024 and spans 31 provincial-level, 365 prefecture-level and 2,672 county-level administrative units in China. The released variables include case_id, case_origination_year, case_code_ori, involved_case_code, land_type, land_scale, province, prefecture and county. The repository also includes validation files for a random sample of 1,000 judicial documents. The validation files contain pipeline-derived values and manual reference annotations for key fields used in the associated Data Descriptor manuscript. The dataset is organized at the document level. Multiple judicial documents belonging to the same dispute can be linked using case_id. Users conducting case-level analyses should deduplicate or aggregate records by case_id. The released dataset contains structured variables derived from judicial documents and does not include the full text of judicial documents from China Judgments Online.