In response to pressure to reach the Millennium Development Goal of improved sanitation access, the Ethiopian government has developed an ambitious plan to achieve 100 per cent access to pit latrines by 2012. The plans to achieve this target rely upon the assumption that universal access to pit latrines will lead to improved health outcomes. Using the Young Lives pro-poor longitudinal data of Ethiopian children, this research uses propensity score matching to test this assumption. Children who experienced a change from no toilet to a household pit latrine between rounds of data collection were compared to those who continue to use a forest/field. The findings show that there is no significant difference between groups in terms of health outcomes and that a pit latrine does not necessarily signal improved methods of waste disposal. Individual and group interviews conducted by Young Lives suggest that poor infrastructure and care for pit latrines deter children from using such facilities and promote a preference for the use of other methods of waste disposal. Policymakers should note that simply increasing access to pit latrines will not necessarily promote better health outcomes, especially when 'improved' sanitation appears to be less clean than other available options.