The Department of Defense (DOD) has only had three completed warranty programs, and data on current programs will not be complete until the early 1980s. This paper explores the reliability improvement warranty (RIW) concept by evaluating commercial analogs, past DOD warranty experience, and the current tri-service RIW trial. Examination of the completed programs does not show conclusive evidence that warranties were a major factor in observed improvement, but certain observations regarding reliability were apparent: (1) Modification after testing is needed to take advantage of field experience and can be promoted without a warranty. (2) Scheduling should be flexible to allow incorporation of test data in subsequent development and production. (3) The contractor should be involved in initial overhaul and repair. (4) RIWs should not be applied to programs subject to extreme quantity or utilization uncertainty. The authors point out that current RIW trials have terms and penalties that vary widely and diminish the likelihood that they will yield conclusive evidence on the value of warranties.