OBJECTIVES: Effectiveness of radiosurgery for head and spinal neoplasms is established. The aim of this study was to systematically review the clinical literature of CyberKnife for people with selected primary and secondary soft tissue lesions. METHODS: A systematic search was conducted for best available clinical data for lung metastasis, primary and secondary liver cancer, locally advanced pancreatic cancer population treated with CyberKnife radiation. Searching using Medline, EMBASE, Cochrane Library took place in September 2013. RESULTS: Only one relevant comparative clinical study (matched-pair analysis) met the inclusion criteria, assessing effectiveness and safety of stereotactic radiosurgery and radiofrequency ablation for colorectal liver metastasis. For other neoplasms single-arm studies were found. Compared to RFA CyberKnife for liver metastasis was significantly better in median local disease free survival, which was 34.4 months vs. 6.0 months, (p<0.001). 1 and 2-year local control rates also favored CK (85.0% vs. 65.0% and 80.0% vs. 61.0%, respectively) but the difference wasn’t significant. However, trend for better OS was found with RFA (34.4 vs. 52.3 months). For lung metastasis, treatment with CK resulted in 24.0-62.0% complete or partial response, 38.0-76.0% patients stabilized. In primary liver tumors OR (CR + PR) was observed in 63.0-86.0% patients, 0.0-29.0% stabilized, median PFS reached 10.0-15.8 months. Inconsistent results were seen in locally advanced pancreatic cancer population. In one study 92.0% responded or stabilized but in other only 1 patient of 77 had PR. Median OS was 6.4-10.3 months. All studies reported mostly mild adverse events after CK. Serious AE were rare. CONCLUSIONS: There is limited quality evidence on the effectiveness and safety of robotic image-guided stereotactic radiosurgery in patients with soft tissue neoplasms. Available studies are highly heterogenic in methods, patients characteristics and outcomes but suggest that CyberKnife may be beneficial in local tumor control. There is a need of well-designed comparative studies.