This paper investigates the process of how actions of community groups are initiated and carried out. It also identifies key characteristics of participation in community action projects. Social actions in two communities in North Carolina are studied from a 10-year period. The unit of analysis is the individual's participation. Interviews are carried out with community leaders and 30 key informants identified as central participants in actions between 1970 and 1980. When descriptions of empirical case studies of individual participation failed to fit a five-stage model, Charles Hoffer's three-stage framework of community action was used to further analyze individual participation in eleven community projects. The degree of participation overlap is examined based on Warren's conception of "horizontal linkages." The findings contradict popular perceptions of community development action projects as either involving large segments of the community or being led by a few highly active persons. Very little overlap in core group participation was found between projects. However, the hypothesis that community projects rely significantly on the internal resources of voluntary organizations, especially civic clubs, receives clear support. There were no evident advantages in participation rates in the larger and faster growing community, but the participants were younger, more often males from the business community, and "newcomers" and generally were less dependent on external financial support.