The first “R” from animal research ethics prescribes the replacement of animal experiments with animal-free alternatives. However, the question of when an animal-free method qualifies as an alternative to animal experiments remains unresolved.

Drawing lessons from another debate in which the word “alternative” is central, the ethical debate on alternatives to germline genome editing, this paper develops a general account of when something qualifies as an alternative to something. It proposes three ethically significant conditions that technique, method, or approach X must meet to qualify as an alternative to Y: (1) X must address the same problem as Y, under an appropriate description of that problem; (2) X must have a reasonable chance of success, compared to Y, in solving the problem; and (3) X must not be ethically unacceptable as a solution. If X meets all these conditions, its relative advantages and disadvantages determine whether it is preferable, indifferent, or dispreferable as an alternative to Y.

This account is then applied to the question of whether animal-free research methods qualify as alternatives to animal research. Doing so breaks down the debate around this question into more focused (ethical and other) issues and illustrates the potential of the account.