A physiological adjuvant to neuroplasticity cannot put new knowledge into the brain - it can only augment behavioral rehabilitation, which does replace knowledge in the brain lost due to injury. Clinical trials of adjuvant agents should therefore incorporate behavioral engines: behavioral therapies that have a potent effect and minimize error variance due to variability in subject attributes and the administration of the therapy. We consider in some detail the attributes of behavioral engines with a particular focus on constraint induced movement therapy. Human trials of physiological adjuvants to neuroplasticity are predicated on discoveries from animal research. Because human trials are expensive and carry risk of injury, it behooves us to lay a comprehensive foundation for these trials with animal studies (also incorporating behavioral engines) and to carefully mesh animal and clinical work. We review in detail our own and others' work involving human subjects with brain lesions and animal models of brain injury using donepezil, a central acetyl cholinesterase inhibitor, d-cycloserine, a partial agonist at the NMDA receptor glycine site, d-amphetamine, and methylphenidate in an effort to elucidate the strengths and weakness of animal and human neurorehabilitation science with respect to this mesh between animal and human subject research. We consider modifications in current approaches that might more efficiently advance neurorehabilitation science.