I found The Illustrated BMI Categories Project off the New York Times health blog. Basically, in terms of what people intuitively think of as “fat,” BMI is a load of bullcrap. Some highlights:
“Overweight”…and gets the Nelson & Karen stamp of SEXY approval!
“Overweight”…and a frickin’ triathlete.
“Nearly obese”…and beautiful. :)
When I started the Weight Watchers deal two months ago or so, I set my goal weight at 160 lbs. The only time I’ve ever been 160 was before I finished puberty, but that’s how skinny I’d need to be to be “normal” according to the BMI calculator I consulted. Now I’m wondering if that’s at all feasible, given how arbitrary BMI appears to be. In terms of health, sure, the medical establishment uses BMI as their independent variable, and there are correlations between high BMI and diabetes, heart disease, etc. But other factors, like abdominal fat, seem to correlate better–so why use BMI, given its flaws? Especially given other studies showing that overweight women survive heart attacks *more* often than “normal” women do. Who knows where the health benefits and the risks intersect with regard to BMI? It seems like a mostly confusing and pointless standard.
The New York Times has its own BMI calculator, but it includes other factors like abdominal fat, physical inactivity, smoking, high blood pressure, and so forth. I still get a BMI of 28.7 (fail), but their calculator says that “if you are overweight, but do not have a high waist measurement and have fewer than two risk factors, you may need to prevent further weight gain rather than lose weight.” Well, as far as I know, I don’t have any other risk factors. Heck, my blood pressure is low. So hmm.
I’m still keeping with the Weight Watchers thing, and trying to exercise more. That’s a good idea no matter what. But maybe I don’t have to freak out if I never quite drop the 30 lbs…

The BMI obviously isn’t the best measure of health. I’m 5’10″ and flunctuate between 140lb and 160lb, putting me in the normal range. But, when I get above 155, I start considering myself too heavy.
I would argue that the reason it is so widely used compared to other better measures is that height and weight are much easier to measure (and height is fairly constant). It is unrealistic to set goals against a measure that requires a doctor consultation to observe and most people are not willing to buy specialized equipment. To use BMI, you only need a $10 scale.
BTW, I think WeightWatchers is not worth the money. You pay them to help you lose weight when you could do so on your own by eating healthy foods and excersizing. The real product that WeightWatchers (and all diets) sells is the added guilt cost associated with not adhering to the program. This is magnified if you diet in a group because the group increases the social cost of breaking the rules. These guilt and social costs provide an incentives for people to do things that contribute to weight loss. I can make myself feel guilty all on my own. :)
(Psst–I’m not paying for it, my mom gave me the materials from when she was on it. Means I don’t have the weekly meetings as an incentive, but I’m fine with that.)
I definitely think WW isn’t a cure-all and people should do whatever works for them. But I found that I needed *some* sort of structure–part because when I started I didn’t have a good sense of the caloric value of some foods and part because it makes it harder to say “well, just this once” when you’re keeping track of what you eat. I did modify the program in some ways, though. For one, I made all fruits and veggies count for zero in exchange for making my daily point allotment a bit lower–you’re never gonna get fat eating apples and tangerines and I didn’t want to have any disincentive to eating them.
When I’m living with Nelson, though, I tend to lose weight without thinking about it just because there’s no all-you-can-eat dining hall and he only buys healthy groceries. Then a structured diet is less necessary.