Research in the 1970s based on observational data provided evidence consistent with predictions from economic theory that paying unemployment insurance (UI) benefits to involuntarily jobless workers prolongs unemployment. However, some scholars also reported estimates that the additional time spent in subsidized job search was productive. That is, UI receipt tended to raise reemployment wages after work search among the unemployed. A series of field experiments in the 1980s investigated positive incentives to overcome the work disincentive effects of UI. These were followed by experiments in the 1990s that evaluated the effects of restrictions on UI eligibility through stronger work search requirements and alternative uses of UI. The new century has seen some related field experiments in employment policy, and reexamination of the earlier experimental results. This paper reviews the experimental evidence and considers it in the context of the current federal-state UI system.