Although a majority of women had heard of malaria in this survey, only a small proportion knew the cause, signs or symptoms, and preventive measures. We show that women belonging to the poorest wealth quintile, with no formal education, and living in rural EAs had lower levels of malaria knowledge 
. These findings are similar to previous smaller studies in Ethiopia, which noted limited malaria knowledge amongst women 
and strong socioeconomic disparity with malaria knowledge (e.g. 96% of respondents in the highest SES quintile had heard of nets versus to 35% in the lowest quintile) 
. These findings contrast with those reported in the Zambia MIS, where almost 100% of the women had heard of malaria, and knowledge disparities amongst wealth quintiles and education levels were less striking 
Women most often received their malaria information from government clinics/hospitals and from friends and family, and least often from print media. This is consistent with the very low rates of literacy reported in Ethiopia. Targeting information, education, communication/behavior change communication (IEC/BCC) efforts to women attending government clinics and hospitals could be an effective way to reach women and change their (and their household's) behavior at a time of high perceived susceptibility 
. With the high illiteracy rates and lack of formal education, less emphasis should be placed on print media as a delivery mechanism.
Further attention is also needed on the content and types of malaria messages delivered. Less than half of the women and mothers knew fever as a symptom of malaria, or knew any danger signs of malaria. Noting the limited emphasis on fever treatment seeking in IEC activities, it is not surprising that rates of treatment seeking are unacceptably low. Increasing the quantity and improving the quality of this messaging could have a large impact on children receiving effective ACTs. Furthermore, receiving messages at home was shown to be significantly associated with improved ITN ownership and use. The 30,000 community-level health extension workers in Ethiopia represent a great opportunity to deliver targeted health messages in people's homes.
We show that ITN ownership was associated with personal or household factors, including women's malaria knowledge, residence in an EA with an altitude below 2000 m and whether the household had received other malaria interventions such as IRS. Although significant malaria transmission and net ownership still occur above 2000 m 
, malaria risk is higher below 2000 m and the increased knowledge may reflect targeting of net distribution and IEC to these areas. Malaria knowledge was associated with increased ownership of ITNs, but we cannot determine in this cross-sectional study whether this is because women had higher knowledge from exposure to more malaria interventions including ITN distribution and IRS campaigns, or because women with greater knowledge were more likely to seek out and accept malaria interventions. This important issue should be further explored through qualitative studies in the future.
The first step and most important factor driving ITN use is ITN ownership. Restricting the analysis to households that own at least one ITN, correct women's knowledge of malaria was associated with both increased ITN use for themselves and their children U5. In such households, increasing number of ITNs was associated with both women's personal use of an ITN and use by their children U5. The ratio of increasing ITNs to household size was explored across 15 standardized national surveys in Africa and generally found to be a significant factor for children U5′s ITN use 
. A cluster randomized trial planning to evaluate the effect of training on LLIN use after LLINs were distributed in south-west Ethiopia, noted higher use with good malaria knowledge and female heads of household at baseline in some villages 
. Improving women's malaria knowledge and increasing the number of ITNs in the household, both modifiable factors, could drive ITN use for women and children U5 once the important initial hurdle of ITN possession has been overcome.
Although recent studies, both population- and health facility-based, have suggested tremendous reductions in malaria morbidity in Ethiopia 
, our survey highlighted the unacceptably low proportion of children U5 accessing any treatment for fever, or let alone prompt and effective treatment 
. Malaria knowledge was not independently associated with seeking care for fever, except for a subset of women with no prior education. For mothers of children U5, the major significant factors in determining care seeking behavior were having attended school, living in an urban area with likely improved access to care, and belonging to the highest wealth quintiles. Those living at altitudes below 2000 m and in households without an ITN were more likely to seek care, which could be attributed to varying perceptions of risk. Studies conducted in central rural Ethiopia prior to universal access to ACTs showed that only 13% 
and 28% 
of children U5 received any form of treatment within 24 hours. Early treatment seeking was more frequently reported by those who accessed home treatment and community health workers 
. The recent deployment of over 30,000 health extension workers should result in large improvements in access to and use of community-based prompt and effective treatment.
Overall, women and more specifically mothers of children U5 reported very low levels of school attendance and literacy. Although women without formal schooling were receptive to malaria messages and were able to access the public health intervention, school attendance was an independent factor associated with women sleeping under an ITN, children U5 sleeping under an ITN, and fever treatment seeking. A previous study in rural Ethiopia found that literacy was a significant factor in women believing that malaria was preventable 
. Furthermore, women's education often emerges as a key element in a strategy to improve overall child health. Formal education improves child health through directly teaching health knowledge to future mothers 
. This is consistent with our study that malaria knowledge independent of school attendance appeared to improve malaria health-related behaviors. However, broader goals of economic development and improving access to education, especially for women remain a significant and modifiable social determinant of health-related behavior and can only aid the malaria control efforts.
Although knowledge is one aspect of a complex interplay of factors, it is an important prerequisite for instigating behavior change and could likely inform attitudes about malaria health-related behaviors. With the global efforts to achieve universal coverage by 2010 
, programs should not only focus on delivery of goods, but packaging that with effective IEC/BCC messages. Lastly, a strategy based on a girl- or woman-centered approach, which is advocated by the Lantos-Hyde U.S. Government Malaria Strategy 2009–2014 
, can be an effective way of increasing the emphasis on girl's education and delivering health messages to this receptive group.