Active for Life((R)) (AFL) was a large (n = 8159) translational initiative to increase physical activity (PA) in midlife and older adults. Translational research calls for a shift in emphasis from just understanding what works (efficacy) to also understanding how it works in more 'real world' settings. This article describes the process evaluation design and findings, discuss how these findings were used to better understand the translational process and provide a set of process evaluation recommendations with community-based translational research. AFL community organizations across the United States implemented one of two evidence-based PA programs (Active Living Every Day-The Cooper Institute; Human Kinetics Inc. or Active Choices-Stanford University). Both programs were based on the transtheoretical model and social cognitive theory. Overall, the process evaluation revealed high-dose delivery and implementation fidelity by quite varied community organizations serving diverse adult populations. Findings reveal most variation occurred for program elements requiring more participant engagement. Additionally, the results show how a collaborative process allowed the organizations to 'fit' the programs to their specific participant base while maintaining fidelity to essential program elements.