The Washington Reemployment Bonus Experiment (WREB) was composed of 6 different "treatments" involving offers of bonuses to over 12,000 unemployed workers filing initial claims for unemployment insurance (UI) benefits. The 6 treatments were represented by all combinations of 3 different "bonus amounts" calculated as either 2, 4, or 6 times the claimant's weekly benefit amoung and 2 different "qualification periods," calculated as either 20 or 40 percent of a claimant's maximum entitled duration of benefits. The experiment sought to identify an optimally cost-effective combination of bonus amount and qualification period. A control group was not offered a bonus. The strongest response to WREB bonus offers was exhibited by those who received high bonus amount-long qualification period offers. For this treatment there was a statistically significant reduction in weeks of insured unemployment of 0.76 weeks over the benefit year and an average reduction of $138 (6.7 percent) in benefit payments over the benefit year. The weakest effect was observed in response to the low bonus amount-short duration offer. On average, this group reduced weeks of insured unemployment by only 0.05. Overall, the estimated mean impact across the six treatment groups was a statistically significant 0.40 week reduction in insured unemployment over the benefit year and a reduction in unemployment compensation paid of $63 over the benefit year. (Appendixes contain 17 references and 21 data tables.) (YLB)