The paper at hand deals with issues arising from inflated values of safety belt use. The paper addresses and evaluates data leading to strong evidence that safety belt use rates, as reported in crash records, are inflated irrespective of reasons. Equally strong evidence is found that belt use inflation introduces significant distortions in the estimation of casualty rates, and of the effectiveness of various restraints. The cited effects are evidenced both analytically and by reviewing crash records. No fully satisfactory remedy is yet at hand for controlling the cited distortions. However, a heuristically determined blanket deflation appears to rectify most of the ill effects of belt use inflation. It also appears that blanket remedies for belt use inflation, intended to cure inflated values of restraint effectiveness, may introduce adverse side effects on the cited effectiveness. Specifically, it appears that an appropriate deflation approach may be a function of restraint type, calendar year, vehicle class and model year, and several other vehicle and occupant attributes. The entire study presented here is based on the crash records of NHTSA's National Accident Sampling System in 1988 to 1996, and NHTSA's National Occupant Protection Use Survey in 1994 and 1996.