This study examined informal controls on smoking (i.e. rules about smoking that individuals impose on one another) in Winnipeg households. A representative sample of 521 individuals was interviewed. Smoking was found to be restricted in 13% of households and prohibited in another 11%. In one-person households, annoyance with smoking was the best predictor of informal controls. In multiple-person households, the best predictors of informal controls were: having a preference for social relationships with non-smokers, believing that second-hand smoke is harmful, and having few friends who smoked. For married/cohabitating respondents, having a spouse/partner who smoked reduced the likelihood of informal controls. The finding that nearly 90% of respondents agreed that environmental tobacco smoke (ETS) was harmful, but only 24% said that smoking was controlled in their residence, indicates that there is a need to investigate the situational variables which can impede the adoption of informal residential smoking norms.