Background: This research was conducted within two Welsh bordering Local Authorities (LAs) across two Educational Psychology Services (EPSs) that changed their service delivery model to consultation following joint training. The Principal Educational Psychologist (PEP) sought to evaluate the EPSs and the services they deliver, but previous methods, including sending questionnaires to head teachers and parents, were not considered robust enough or to have sufficient depth. The PEP therefore commissioned this piece of research to source a clear and useful measure to evaluate consultation. Participants: Across the two LAs, the PEP, ten generic Educational Psychologists (EPs) and one Senior Educational Psychologist (SEP) participated in the study.Methods: This study used the Research and Development in Organisations (RADIO) model of Action Research (AR). A literature review was initially conducted to identify potential evaluation models. Thereafter, within the AR phases, different data were collected and analysed with stakeholders to ascertain EPs' use of evaluation models and their preferences and perspectives as practitioners. Data collection incorporated both focus groups and questionnaires, which generated quantitative and qualitative data which were analysed through a variety of methods, including content analysis, thematic analysis and questionnaire analysis software. Findings: Within the AR design, data gathered during the earlier phases of the research were used to guide next steps in the research process. This led to the identification of two preferred models: The Constructionist Model of Informed Reasoned Action (COMOIRA) and Appreciative Inquiry (AI). These were then piloted across the EPSs and each EP completed a post-pilot questionnaire. In the final phase of the research two focus groups were held to consider the usefulness of the COMOIRA and AI models as service evaluation tools. The data generated a number of themes relating to accountability, applying psychology, change, evaluation and strengths; and raised issues relating to future implications. Conclusion: The AR design of the research facilitated a collaborative approach for shared decision making around the development of an EPS evaluation framework. The study identified positive and valuable aspects with regards to both the COMOIRA and AI models when these were implemented to evaluate consultation. Both models have the potential to assess aspects of the consultation process and could conceivably contribute towards providing outcomes which demonstrate accountability to employers and service users. Additionally, professional practice models could be used to evaluate wider aspects of EP practice and be of broader benefit to EPSs.