One of the popular realist responses to the pessimistic meta-induction (PMI) is the ‘selective’ move, where a realist only commits to the ‘working posits’ of a successful theory, and withholds commitment to ‘idle posits’. Antirealists often criticise selective realists for not being able to articulate exactly what is meant by ‘working’ and/or not being able to identify the working posits except in hindsight. This paper aims to establish two results: (i) sometimes a proposition is, in an important sense, ‘doing work’, and yet does not warrant realist commitment, and (ii) the realist will be able to respond to PMI-style historical challenges if she can merely show that certain selected posits do not require realist commitment (ignoring the question of which posits do). These two results act to significantly adjust the dialectic vis-à-vis PMI-style challenges to selective realism.