Performance evaluation is a central governance mechanism in organizations because it shapes what becomes visible, legitimate, and valued. In social work, this poses particular challenges, as practice relies on professional judgement, relational engagement, and context-sensitive interventions whose effects are cumulative and preventive rather than immediately measurable. Indicator-based performance regimes associated with New Public Management (NPM) have intensified these challenges by privileging standardized and quantifiable outputs as dominant and ostensibly neutral forms of knowledge. In Nordic reform debates, trust-based governance (TBG) has been proposed as a response to strengthen the legitimacy of professional discretion and dialogue and to reduce micro-management control. However, little research has examined how such ambitions are translated into core governance instruments, particularly performance evaluation. This exploratory qualitative case study investigates how performance evaluation is articulated when front-line social workers are invited to define what should be evaluated in a Swedish municipality undergoing TBG-oriented reform. Data were generated through four participatory workshops (N = 89), in which practitioners first articulated their practice logic and then developed proposals for evaluation approaches aligned with that logic. The findings show that practitioners did not reject measurement but reframed evaluation as an interpretive and dialogical process. Indicators were regarded as useful but insufficient without contextual interpretation, qualitative reflection, and structured dialogue. The study demonstrates that operationalizing TBG depends on reshaping evaluative practices so that accountability can coexist with professional judgement and relational forms of knowledge.