The purpose of the current project is to explore the use of propensity scores to estimate the effects of interventions within randomized control trials, accounting for varying levels of implementation or fidelity. This work extends that of Jo and Stuart (2009) to settings with multiple or continuous measures of implementation. Rather than focus purely on a binary measure of compliance or non-compliance in the treatment group, the study looks at levels of compliance to treatment using a continuous measure of participation or fidelity. The methods are investigated using simulated data as well as data from the Infant Health and Development Program (IHDP), an eight-site randomized trial that targeted low-birth-weight premature infants, and offered comprehensive, quality early intervention care with the goal of reducing the developmental and health problems of low-birth-weight infants (IHDP, 1990).