Systemic inflammation is associated with cardiovascular disease, especially the end stages of heart failure (HF). The authors sought to determine by systematic review whether physical therapy reduces serum levels of pro-inflammatory cytokines in HF patients. Potential studies were identified from a systematic search of Medline (Ovid) (1950-October 2010), Embase.com (1974-current), Cochrane Central Register of Controlled Trials, CINAHL (1981-current), and Web of Science (2000-current). The search strategy included a mix of MeSH and free text terms for key concepts. A total of 11 studies were included, with 4 reporting a post-training reduction in tumor necrosis factor α (TNF-α), of which 3 used traditional aerobic or resistance exercise and 1 used functional electrical stimulation. Reduced post-training serum levels of interleukin 6 were reported by one exercise study. The one study that employed combined resistance and aerobic training only showed soluble receptors TNF-α1 and TNF-α2 to be lower. The one study of respiratory muscle training and two studies that employed electrical stimulation had limited effect on cytokines and peak maximal oxygen uptake. With the exception of one study, those therapies that employed ≥ 5 sessions per week lowered serum TNF-α. Our review data suggest that physical therapies employing ≥ 5 sessions per week are most likely to reduce serum levels of TNF-α in HF patients. No isolated study of resistance training has yet examined the relationship between muscle wasting, glycogen depletion, and the ability of exercise training to reverse both of these conditions in HF patients.