Despite criticisms of the participant observation method, most field workers seem satisfied that their research accurately depicts social reality and the results are scientifically valid. Research on career criminals and criminal groups presents special methodological barriers, such as the secretive nature of criminal syndicates and reliance on secondary and tertiary accounts of events. A review of recent studies using participant observation techniques reveals the methodological influence of Ned Polsky. His views clashed with the traditional belief that criminals are disinclined to provide information about themselves or their illegal activities. He also argued that defined and controlled techniques of observation were something of a fetish that impeded rather than aided research. Other social scientists have commented that discussions of qualitative research ignore the fact that role relationships among researchers and subjects are bound to change. Polsky maintained that participant observation of criminals involved getting thoroughly familiar with their frame of reference, language, and interests. Once entry has been accomplished, sustenance of the research situation requires the clear separation of the researcher from the criminal activity and a lack of pretense regarding the purpose of the investigation. Cultivating rapport triggers a snowball effect in which a network of informants is likely to emerge. Data can be validated through this network as well as by law enforcement contacts. Research on organized crime intensifies both methodological and ethical problems, and investigators cannot expect conclusive results no matter how ingeniously they manipulate Polsky's strategies. The paper contains 22 references.