This talk explores the possibility of incomparability.  I will first focus on a challenge to the small-improvement argument for incomparability and then turn to a seemingly more promising variation on it: the huge-improvement argument for incomparability.  Ultimately, there seems to be a way around this argument too, but reflection on the argument and on getting around it is revealing.  It suggests that if there are any cases of incomparability, the really interesting phenomenon in such cases is not between the options but within the options, or at least within one of them: more specifically, it suggests that cases of incomparability, if there are any, are cases in which at least one of the options is resistant to classification as positive, negative, or fairly neutral.  The answer to the question of whether every (contextualized) option can be classified as positive, negative, or fairly neutral is not obvious.  I will put forward a candidate case of an option that cannot be so classified, and a supporting argument, namely the huge-improvement argument for resistance to overall evaluative classification.  This argument seems to provide a promising basis for the possibility of incomparability, but consideration of an important complication suggests that there remains a gap between the possibility of options that are resistant to classification as positive, negative, or fairly neutral, and the possibility of incomparability.  In the end, what matters is that we recognize that, whether or not options can be strictly incomparable, there is room for cases beyond not just classic cases of precise comparability but even beyond cases of coarse comparability involving positive options, negative options, or fairly neutral options.  In the relevant further cases, at least one option is not positive, negative, neutral, or even fairly neutral.  To the extent that comparability is revealed as applicable in such cases, skepticism about incomparability can persist, but its significance is reduced by the revelation of how little comparability requires.