You ask a good question that I have wondered about myself. The classic examples of immoral work in science are Nazi experiments on human physiology and the Tuskegee syphilis study. Neither were up to current methodological standards, but both were OK science for their time. In a way it would be more convenient if these cases would be bad science as well as immoral science, because then no questions need be asked about whether it is permissable to use the results. Perhaps it is difficult to acknowledge that science can be used successfully in ways that are immoral. But I think we learned this lesson with Hiroshima and Nagasaki.
Read another response by Miriam Solomon
Do immoral methods in science always produce false results? I've heard this kind of claim made in relation to psychological experiments in which subjects are initially lied to. It doesn't seem intuitive. Why do people say this?