Objectives: Health care utilization in Burkina Faso is far below the average level in OECD countries (0.15 and eight contacts per capita and year respectively). Worse still, there is a pronounced income elasticity of demand. Hence, much of the potential benefit of effective drugs remains elusive for most of the sick poor. This paper reports on the enrollment of the poorest quintile in CBI and the change in health care utilization of the insured households. Further more, the effect of a targeted subsidy on enrollment of the poorest 5th of household was assessed. Method: Community-based insurance has been offered to a district in Burkina Faso, comprising 74,000 people who lived in 53 villages and the district capital of Nouna since 2004. Community self assessment of poverty was used in 2007 to identify the poorest quintile of households who were subsequently offered insurance at half the usual premium rate. Results: Overall enrollment in health insurance was 5.2% in 2006 and 8.3% in 2007. However, only 1.1% of the poorest quintile of household chose to enroll. In 2007, with the targeted subsidy on enrollment of poorest households, this rate reached 11.1%. Once enrolled, households increased their utilization of health care substantially (from 0.15 to 0.64 contacts per capita and year). This increase in utilization given enrollment held also for the poorest quintile. Conclusion: The authors discuss the covariates influencing household decisions to enroll and to use health care subsequently. They stress the need to enhance the access of the poor to drugs and health care in general through the combined effect of community-based insurance and targeted subsidies to the poor.