Background: Mortality in African children is unacceptably high. Providing free health care to young children has been hypothesized to improve access to health care and lead to better health outcomes. However, there have been only few methodologically robust studies testing this hypothesis. This study examined the impact of free healthcare provided to a birth cohort of children born from March, 2005 to June, 2006 in Butajira DSA (Demographic Surveillance Area) through the C-MaMiE project (Child outcomes in relation to Maternal Mental health in Ethiopia), in comparison to three cohorts of user fee children (children born within 2.5 years before and after the intervention and parallel with the intervention). Objective: The aim of this study was to assess the impact of making health care services free at the point of use upon under-fives mortality. Methods: This controlled quasi-experimental study compared intervention cohort, children born in Butajira from March, 2005 to June, 2006 and received free health care provided by the C-MaMiE project, from pre-birth (pregnancy) to 12 months, and from 24 to 60 months, and three comparison cohorts of children born in the same geographical area (2.5 years before, during and after the intervention). The crude and adjusted under-five mortality in the intervention cohort versus the comparison cohorts was evaluated using Cox regression model. Result: Incidence of under-five mortality was 15.7 (12.3-20.0), 98.7 (88.3-110.3), 41.2 (35.2- 48.2) and 39.3 (33.5-46.1) per 1000 person-years of observation children provided free, user fee before intervention, user fee during intervention and user fee after intervention health care services, respectively. The risk of under-five mortality among user fee children (before, during and after the intervention) were 5.87 (4.47-7.72), 2.45 (1.84-3.28), and 2.72 (2.03-3.66) times higher than those provided freely, respectively. Maternal death (AHR=2.10; 95% CI; 1.36- 3.23), rural residence (AHR=0.56; 95% CI; 0.44-0.72) and death of elder child (AHR=1.28; 95% CI; 1.01-1.61) were also found to be independent predictors of under-five mortality. Conclusion and recommendation: There was a slow decrement of early childhood mortality in the study area and was relatively high among user fee children, even compared to the after arm comparison cohort whereby lower mortality is expected, and health policy makers should give more emphasis on abolition of user fee health care services as it is one of the methods which significantly reduce under five mortality