Background Sustainability of evidence-based health promotion interventions has received increased research attention in recent years. This paper reports sustainability data from Project HEAL (Health through Early Awareness and Learning) a cancer communication implementation trial about early detection, based in African American churches. In this paper, we used a framework by Scheirer and Dearing (Am J Publ Health 101:2059-2067, 2011) to evaluate multiple dimensions of sustainability from Project HEAL. Methods We examined the following dimensions of sustainability: (a) continued benefits for intervention recipients, (b) continuation of intervention activities, c) maintaining community partnerships, (d) changes in organizational policies or structures, (e) sustained attention to the underlying issues, (f) diffusion to additional sites, or even (g) unplanned consequences of the intervention. Project HEAL provided a three-workshop cancer educational series delivered by trained lay peer community health advisors (CHAs) in their churches. Multiple sources of sustainability were collected at 12 and 24 months after the intervention that reflect several levels of analysis: participant surveys; interviews with CHAs; records from the project’s management database; and open-ended comments from CHAs, staff, and community partners. Results Outcomes differ for each dimension of sustainability. For continued benefit, 39 and 37% of the initial 375 church members attended the 12- and 24-month follow-up workshops, respectively. Most participants reported sharing the information from Project HEAL with family or friends (92% at 12 months; 87% at 24 months). For continuation of intervention activities, some CHAs reported that the churches held at least one additional cancer educational workshop (33% at 12 months; 24% at 24 months), but many more CHAs reported subsequent health activities in their churches (71% at 12 months; 52% at 24 months). No church replicated the original series of three workshops. Additional data confirm the maintenance of community partnerships, some changes in church health policies, and continued attention to health issues by churches and CHAs. Conclusions The multiple dimensions of sustainability require different data sources and levels of analysis and show varied sustainability outcomes in this project. The findings reinforce the dynamic nature of evidence-based health interventions in community contexts. Electronic supplementary material The online version of this article (doi:10.1186/s13012-017-0576-x) contains supplementary material, which is available to authorized users.