The New York State Assembly is poised to consider a bill requiring H.I.V. testing of suspects who have been indicted for rape. Governor Elliott Spitzer and a majority of the Assembly support the legislation. However, it is nonetheless controversial and may therefore stall before coming up for a vote. The question of whether such legislation properly balances the interests of rape victims and rape suspects is a difficult one.

Those who support the bill say that a rape victim should have as much information as possible in deciding whether to begin taking the combination of H.I.V. medications known as the "cocktail" after a sexual assault. These medicines play a crucial role in slowing down the multiplication of the virus and in strengthening a person's immunity, but they have many unpleasant and significant side effects. To ensure compliance, it is therefore useful for a patient to know that she has actually been exposed to the virus. The reason the victim cannot simply take an H.I.V. test herself right after an assault is that it takes about twelve weeks for an infected person's blood to become "H.I.V.-positive" for purposes of an antibody screen.

Opponents of such testing respond along two different dimensions. The first argument emphasizes the invasion of a suspect's privacy involved in testing his blood. Because a suspect indicted for rape is "innocent until proven guilty," he should not have to take a test the premise of which is that he did in fact rape the complainant. The second argument points out that such testing does not truly further the health interests of the rape victim. A victim who relies on a suspect's negative H.I.V. test to forego the cocktail could actually be making a life-threatening mistake: the victim could be mistaken about the identity of her attacker (and the actual rapist could be H.I.V.-positive) or the suspect, if he is guilty, could have only recently contracted the virus, in which case his negative result would be meaningless.

On the question of suspect privacy, I am quite sympathetic. A person who receives a positive H.I.V. test confronts the possibility of discrimination from many quarters, including the insurance industry and employers. The fact that many testing centers offer confidentiality of results is quite revealing, in this regard, and it is not necessarily realistic (or even fair) to expect a rape complainant to keep such information secret. An indictment does signify a grand jury's belief in "probable cause" to prosecute the defendant for rape. Still, the prosecutor runs the grand jury process, and it is therefore vulnerable to errors that a competent defense attorney could later expose at trial. Furthermore, because most of the rapes that are prosecuted (though not the majority of rapes that occur) are "stranger rapes," there is a real possibility that the victim unwittingly identified the wrong person. It would be quite unfortunate if a suspect were first wrongly accused of rape and then publicly exposed as H.I.V.-positive.

Nonetheless, the notion of "innocent until proven guilty" is primarily a presumption that governs the trial jury's consideration of the evidence. Beyond the conduct of the trial, the law does not treat a person indicted for rape (or murder, etc.) just like all other innocent people in society. For example, many people are held in jail pending trial on the basis of a finding that leaving them at large poses an unacceptable risk to the law-abiding population (and the risk of flight). If such suspects try to escape confinement, moreover, deadly force may be used to stop them. By comparison to jail (where suspects have virtually no privacy) and the possibility of deadly force, an H.I.V. test seems far less invasive.

I am less sympathetic still to arguments denying the utility of such testing. A doctor surely ought to tell her patient that a person indicted for rape could actually be innocent and that even if he is guilty, he might have contracted H.I.V. recently enough to "pass" an H.I.V. test. Such information is part of informed consent. It is nonetheless informative for a rape victim to learn how the person she believes attacked her fares on an H.I.V. test. For one thing, a negative result could give her some peace of mind during a very traumatic time in her life. For another, a negative test -- while, for reasons stated, is not foolproof -- can help her make an intelligent plan for how to address the possibility of infection.

In all, I think I have persuaded myself that a law mandating H.I.V. testing for people indicted for rape would be a net positive development. It does subject a defendant who has yet to be tried (and thus may be innocent) to an invasion of privacy that may well result in stigma and discrimination. Being indicted for rape carries its own stigma too, though. If available information about her indicted assailant could ease the severity of a rape victim's trauma or motivate her to undergo life-preserving drug therapy, it seems worthwhile (though not cost-free).