We introduce "stigma sentiments" as a way to operationalize the cultural conceptions of the mentally ill. Stigma sentiments are the evaluation, potency, and activity (EPA) associated with the cultural category "a mentally ill person." We find consistent support for the validity of the evaluation and potency components as measures of these conceptions. We also demonstrate the validity of EPA ratings of self-identities ("myself as I really am") and reflected appraisals ("myself as others see me") as measures of self-meaning among psychiatric patients. Then we assess hypotheses derived from the modified labeling theory of mental illness: we expect each stigma sentiment to be related positively to the corresponding dimension of self-identities and reflected appraisals among psychiatric patients but unrelated to the corresponding dimension among non-patients. We find support for 13 of the 18 components to these hypotheses. We also find three cross-dimensional results that were not anticipated. The results suggest that the cultural conceptions of the mentally ill do affect the self-meanings of individuals diagnosed with a mental disorder, although the connection is sometimes more complex than a one-to-one relationship between a stigma sentiment and its corresponding dimension of self-meaning. (Contains 6 figures, 6 tables, and 5 footnotes.)