Background Nearly 2.4 million Beijing residents experience intimate partner violence (IPV) annually. Of these 2.4 million, over 800 000 are injured by IPV; more than 300 000 are injured badly enough to require medical attention. Informal social control exerted by neighbours in communities with high levels of family–community integration (like those made up of residents of traditional courtyard house-and-alley Beijing neighbourhoods called ‘Hutongs’) may protect against IPV injury compared with apartment dwellers.

Methods We tested the protective effects of informal social control and Hutong residence in a randomly selected, three-stage cluster sample of Beijing families reporting IPV. Informal social control of IPV (ISC_IPV) was measured using two 7-question Likert scales developed by the first author. Interviewers were given detailed instructions on how to classify neighbourhoods as Hutong-style or not. We used a Sobel test to examine whether the Hutong effect was mediated by informal social control. The initial sample was of 506 families. Analyses were carried out on 113 families who reported any IPV in the last year.

Results Random effects regression models showed that both acts of informal social control and Hutong residence were associated with less IPV injury. However, the protective finding for Hutong residence was not explained by informal social control, collective efficacy, characteristics of the IPV or demographic characteristics of respondents and households.

Conclusions The unique protective association with Hutong residence suggests that the benefits of community life remain insufficiently theorised and understood.