ABSTRACT. Malware analysis is one of the critical domains of cybersecurity, since malware is constantly evolving, new techniques and tactics are developed by adversaries. Several different methods and types of malware detection are present in defensive security specialists’ tool set, but evasion techniques are just as vast and diverse. Following paper tries to thoroughly scrutinize different techniques malware authors utilize to circumvent analysis, and categorize them based on type of defensive methods they are trying to bypass. Research explores common forms of code obfuscation used to dodge detection by automatic and static analysis, Methods for detecting VM, sandbox or debugger during manual dynamic analysis and how recently developed machine learning solutions are applied in defensive context. Lastly, paper concludes what are the main limitations of contemporary solutions and highlights remaining vulnerabilities and blind-spots in modern defensive approaches