In the aftermath of Peter Singer’s ‘Famine, Affluence and Morality’, the argument he put forward received significant criticism, largely on the grounds that it demanded too much of moral agents. Several attempts have been made since to formulate moral principles that adequately express the stringency of our duties of beneficence. Richard Miller proposed one such option, which has several advantages over Singer’s principle. In particular, because it concerns our dispositions rather than operating over every possible occasion for beneficence, it avoids problems of iterative demands. However, I argue that Miller’s principle is inadequate, because 1) it seems too weak, 2) it appears to be ambiguous and 3) it can give unduly harsh verdicts on unlucky moral agents.