Promoting women's employment is not only the need of social and economic development, but also the historical mission of liberating women. This paper uses data from the 1％ Population Sample Survey, taken in Guangdong Province in 2015, to explore how women's marital status, education, and family environment affect the female non-agricultural employment rate (FNAER) on a county scale using a spatial-lag model. The results show that: 1) The female non-agricultural employment rate in counties of Guangdong Prov-ince is low, with more than three-quarters of counties having female non-agricultural employment rate less than 50％. Moreover, the spatial distribution of FNAER is uneven, with the high-value areas concentrated in the southeast and the low-value areas mainly distrib-uted in the central and western parts of Guangdong Province. 2) From the perspective of industry, there are significant spatial differences among women. In the southeast, women are mainly engaged in the secondary industry, while in the central and western regions, women are mainly engaged in the tertiary industry. 3) Women having better skills and more effective support from the elderly can improve the FNAER. Women having lower skills, smaller-scale families, a higher fertility rate, and households with two or more elderly members have a negative effect on the FNAER. 4) Public policies suggest that improving women's education and their family environment, building social welfare facilities, and repairing the family environment will increase the FNAER.