Race is one of the most inflammatory, slippery, maddeningly paradoxical concepts to afflict human consciousness; witness its ugly history. Shamefully, perversions of biology, anthropology and psychology have at various times racially justified colonialism, slavery and disenfranchisement. Medicine's own intersections with concepts of race have tended to be horrible as well: the grotesque Nazi experiments and the notorious Tuskegee studies of syphilis spring to mind.

Looking to change that awful record for the better is the drug BiDil, approved in 2005 to reduce the toll of congestive heart failure specifically among African-Americans. BiDil is not a product of research on the human genome, but researchers in the field of pharmacogenomics are combing our DNA for clues to new therapies—and better ways to match them to appropriate groups of patients. Because investigators will inevitably keep looking for correlations with racial groups, they will keep finding them. Thus, the era of race-based medicine dawns.

The validity of race-based medicine ultimately depends on whether race is biologically meaningful and what that meaning is. The answers get murky. People of different races are genetically different—but so, too, are those of the same race. Differences in genetic traits, such as pigmentation, help to mark individuals' membership in races, but the genes involved vary so much within racial groups that effectively none of them reliably distinguishes the races on its own. Constellations of DNA sequences or genetic traits may be more common among certain ethnicities. But those populations often correspond only loosely to the broad racial categories in common use, which dilutes whatever physiological significance they might have had (Nigerians, Ethiopians and Jamaicans have distinct ancestries, but all get lumped together as black). Moreover, people routinely have forebears of different races, which makes a mockery of the silly rules for assigning them to racial groups. To a first approximation, then, race is biologically meaningless.

On the other hand, because societies do sort individuals into racial categories and treat them accordingly, race is far from meaningless with respect to sociology and public health. Bad environments arising from segregation and prejudice can be systematically hurtful. A society's history of racism can thereby lock medical risks into ethnic groups even in the absence of clear genetic differences.

Given the high rate of cardiovascular disease among African-Americans, any real remedy is one for which to be grateful. As law professor Jonathan Kahn recounts in “Race in a Bottle” (starting on page 40), however, BiDil's race-specific benefit was rather desperately teased out of studies on mixed populations, and it was not very large. Moreover, the strategy of marketing BiDil as a racial drug seems to have been developed to keep the combination of generic compounds in it under patent protection. It is hard not to be cynical about whether the drugmakers put profits ahead of prudence in a sensitive subject.

Even if BiDil is worth the distress of treading on racial sensibilities, future ethnic drugs may not be. As a society, we should know how to weigh the pros and cons. The situation will not always be black and white.