Several characteristics of biotech industry structure follow cyclical patterns. Mergers and acquisitions activity shows cyclical behavior, with peaks from 1988-92 and 1996-97 and a valley from 1993-95. The ratio of large-firm to small-firm field trials, and the Herfindahl-Hirshmann concentration index, move pro-cyclically with M&A activity. This paper develops a formal, dynamic, neo-Schumpeterian model of endogenous R&D and innovation. The model generalizes and extends the literature on biotech industry concentration. For specified parameter values, the out-of-steady-state dynamics are examined, and shown to generate model behavior which is consistent with empirical descriptions of biotech concentration and R&D-activity cycles.