A marketing company conducting a consumer omnibus survey (COS) in April 2000 was paid to include questions about household ownership of mosquito nets. At the time of the survey, which involved 5018 respondents, most and perhaps all of the nets owned by the respondents would have been untreated, as a product for net treatment was then virtually unavailable. Sampling was conducted in all 36 states and the Federal Capital Territory, Abuja. Core questions were asked about socio-demographic and socio-economic characteristics, and commissioned questions were asked about household net ownership, the source of the household's newest net, if any, and the price paid for that net. The data were analysed using bivariate and multivariate methods. Socio-economic status was measured using an asset index. The estimated number of nets was adjusted using information on asset ownership from the national Demographic and Health Survey of 1999. After this adjustment for sampling bias, net coverage and the total number of nets in Nigerian households were estimated to be approximately 9% and 3 million, respectively. The single most important association with net ownership was access to a flit-gun. It appears that COS are a potentially useful source of information for assessing mosquito-net coverage at household level and monitoring changes over time. They have the advantage of being administered frequently, and they are relatively low cost. The present results were comparable with those from other studies conducted at a similar time. The main disadvantage of COS is sampling bias but, as shown here, it is often possible to adjust for this.