Although evaluation is at the core of many of the practices associated with poetry--including teaching, editing, selecting, judging, and even writing--and although there have been involved discussions of the assessment of verse, there has been no empirical investigation of the specific values which, one supposes, lie at the heart of such judgments. Aiming to extend the conversation about the omnipresent but largely ignored issue of evaluation in contemporary American poetry, this study examines how one group of poets valued twelve poems. In pursuit of this axiological insight, the authors developed a method by which seven poets would discuss and reveal evaluative criteria that usually remain unknown and hidden, even to themselves. Once that discussion took place, the authors conducted a systematic analysis to illuminate its evaluative dynamics. The study revealed two very different models by which poems were valued as texts, and it uncovered the powerful role that context plays in assessments. The authors learned some of the unrecognized, multiple, and even contradictory ways in which poems are valued. Although such findings are intrinsically useful, the authors believe that this study is most promising insofar as it presents a technique for further such analyses, an approach that can be used by any group of people--poets, teachers, editors--interested in finding what they value in any kind of poetry. (Contains 2 notes.)