The distinction between evaluative and non evaluative meaning in discourse is problematic. Many expressions may be purely factual in some contexts and evaluative in others. As argued by Hunston (1999:199--201), the criteria for evaluation rely in part on shared assumptions that form part of the message of any texts. Evaluation is, therefore, often expressed in ways that are highly implicit and discourse-dependent. The aim of this paper is to illustrate how detailed manual contrastive linguistic analysis comparing equivalent genres (online property descriptions in this case) produced in two different contexts of culture (England and Italy) may be used to: a) explore the fine variation in the subjective force of evaluative expressions, which might escape automated larger-scale analysis and, thereby b) distinguish with some precision between genre- and culture-specific aspects of evaluation. The theoretical framework for the analysis is derived from Appraisal Theory (Martin and White, 2005) and from the author’s earlier adaptation and application of the framework to the analysis of English property descriptions (Pounds, 2011). The significance of the findings is discussed with particular reference to further research into genre- and culture-specific evaluative patterns.