NHTSA has released their procedure for rollover testing for review. This procedure was established as part of the TREAD act of 2000 - fallout from the Ford Explorer/Firestone tire incidents.

See here for C&D's article

The DOT's press release

The current proposed procedure

The rollover tests themselves are, to me, a moot point. Nothing I drive or plan on driving in the near future will have an issue with rollovers, unless I buy a truck as a tow vehicle. In that case, the trailer controls the rollover resistance anyway, so ultimately NHTSA's rating means little to me personally. The ratings *will* affect SUVs, auto-pickups (pickups driven in a car-like manner), and minivans, though. The likely test will be some variation of the Fishhook maneuver, which is a timed event that maximizes the likelihood of an untripped rollover (worst case scenario testing). I don't believe that any truck/minivan/SUV will pass this test at 50 MPH, especially since a lot of them are borderline for failing the J-turn @ 60 MPH which is a less severe test for untripped rollover.

What does concern me is the addendum to the rollover testing. In order to prevent manufacturers from designing cars to not rollover at the expense of other aspects of handling (safety), NHTSA wants to include a handling grade (A, B, or C) either in addition to or in lieu of the rollover test. In section IX, they propose a batch of testing to determine this grade. The tests they propose include:

- Slowly Increasing Steer maneuver (more commonly referred to as a constant speed steer sweep). This is a very useful test.
- Dropped Throttle in a Turn maneuver. See my comments below on this dandy.
- The Step Steer maneuver. This is another useful test.

There would be 12 scores developed from these 3 tests, each of which is done

Section IX - Handling Tests
Sub-section B - Guiding Principles for NHTSA Handling Testing and Handling Rating
[QUOTE]
There are two perspectives for handling ratings.

Elistan, your second point is a very good one that I hadn't considered. NHTSA's mandate is to increase the public knowledge of how safe their vehicles are in rollover situations; the flip side is that the Insurance agencies could utilize these very ratings to penalize owners of "unsafe" cars that garner a "B" or "C" rating. The flaw is that, once again, these tests show a severe bias towards FWD pass-cars. A good driver in a RWD sports car can be safer though, in that they can avoid 90% of the situations where these "safe" cars get into accidents.

Your first point is also dead on. Understeer is safe. Oversteer isn't, at least not in the hands of most drivers.