The potential contribution of KAP studies to malaria research and control has not received much attention in most Southern African countries. In Swaziland, this is the first study that has been carried out to provide baseline data about malaria related knowledge, attitude and practices at community level prior to the implementation of the malaria elimination strategy.
The results showed that most people had information about malaria. The most important source of information is health facilities. There was little information coming from the preferred source such as the tinkhundlas (traditional community district meetings). Similarly, measures promoted by the Department of Health such as community health workers (CHWs) and rural health motivators (RHMs) generated very little information about malaria. These are interesting results given the fact that malaria in this region is no longer endemic and compares well with studies in endemic countries such as Nigeria . Hearing about malaria is not enough, but should be seen as a foundation through which a whole range of issues about malaria should be understood, for example, malaria transmission, signs and symptoms, prevention and treatment.
In this study, almost all (99.7%) of those who had heard about malaria made correct association between malaria and mosquito bite. These are encouraging results when compared to only 34% of people who made correct association in Zanzibar . Investigations on communities' knowledge of signs and symptoms showed that over 70% of the respondents identified headache, high temperature/fever and chills as the most common ones. This is in line with the observations of most studies in endemic settings [9–11].
The analysis also showed that most respondents seek treatment in the health facilities. Contrary to most sub-Saharan African countries, where treatment is sought mainly in non-public sources [9, 11–13], this difference could probably be attributed to better quality and accessibility of heath facilities in Swaziland compared to the other countries in the continent. Another interesting finding was that the majority (88.1%) of respondents in this study stated that they would seek treatment within 24 hours of onset of malaria symptoms. This exceeds the target defined by the Abuja summit on malaria, which says, *at least 60% of those suffering from malaria should seek treatment within 24 hours of the onset of symptoms' . There was only one case where the respondent had no knowledge, and this was a household head with no education and employment.
Observations regarding preventive measures showed that most respondents (78.1%) believed that malaria is preventable, and mentioned, clinic, spraying and the use of bed nets as key malaria preventive measures. Despite these positive responses a substantial number of them (43.4%) did not take any personal protective measures to guard against malaria infection. This may be due to the fact that most people are dependant on interventions by the Malaria Control Programmes for protection against malaria infection.
The study found that 87.2% of the households had been sprayed during 2006 malaria season. This is in agreement with WHO guidelines on IRS coverage which recommends that it should be more than 80% within the targeted communities . Only a small proportion of respondents were not happy about IRS. Interestingly there was also relatively good bed net coverage (38.8%), especially given the fact that IRS is the mainstay of malaria vector control in Swaziland. A small number of respondents stated that they did not use bed nets because of low mosquito population density and low disease incidence. Given the changing malaria situation in the country continued efforts are needed to emphasise the benefits of operational vector control activities for eliminating localized residual foci of transmission through community health promotion.