BACKGROUND: Understanding variations in women's work participation at lower administrative levels, such as districts, is a missing link in identifying trends, patterns and variation that can offer insights into this long-term stagnation. We link data from the 2019-21 Indian National Family Health Survey and the 2011 Indian Population and Housing Census to generate estimates of women's work within 640 districts of India, and to examine the spatial clustering of women's work across these districts. We examine women's work through three outcome variables, namely, district-level estimates of 1) percentage of women who worked in the past 12 months, 2) percentage of women who were self-employed in the past 12 months, and 3) percentage of women who earned cash in the past 12 months. RESULTS: Diagnostic measures confirm that our model-based estimates are robust enough to provide reliable district-level estimates of women's work in India. Women's work and cash earnings were lowest in the districts of the central, eastern, and northern regions, and highest in the southern region. Self-employment rates for women were generally low in Indian districts, except for districts in Himachal Pradesh and the north-eastern region. CONCLUSIONS: Considerable spatial heterogeneity in women's work has been found across 640 districts of India. Our study demonstrates that estimated percentage of women who worked in the past 12 months, estimated percentage of women who earned cash in the past 12 months and estimated percentage of women who were self-employed in the past 12 months all vary substantially at the district level. Having only state-level estimates may thus be inadequate to inform efforts to remediate low levels of women's work in India. The insights from our current study may help in the formulation and implementation of targeted policies that increase opportunities for women to expand their paid work in India.