Attention has turned to welfare advice as a potential health and social care intervention. However, establishing direct evidence of health impact has proven difficult. This is compounded by the need to understand both the facilitative contexts and mechanisms through which this impact occurs. This study investigated if, how and in which circumstances an intensive advice service had an impact on stress and wellbeing (as precursors to health impacts), for clients attending a branch of Citizens Advice, located in the North East of England. A mixed methods realist evaluation of three intensive advice services offered by Citizens Advice (CA) was operationalised in five phases: (1) Building programme theories (2) refining programme theories (3) Development of a data recording tool (4) Testing programme theories with empirical data (5) Impact interviews. This paper focuses on phase 4. The Warwick Edinburgh Mental Wellbeing Scale (WEMWBS) and Perceived Stress Scale (PSS) were completed by 191 clients, with a 91% follow up rate (data collected: February 2016 to March 2017). Twenty-two CA clients participated in interviews (data collected: October 2015 to November 2016). The PSS indicated a significant decrease in stress from initial consultation to approximately 4-6 weeks post advice from 31.4 to 10.3 (p<0.001) and the WEMWBS showed a significant increase in client wellbeing from a mean of 26.9 to 46.5 (p<0.001). Nine refined programme theories are presented which combine the qualitative and quantitative analysis; they are underpinned by three abstract theories: Capabilities model, The Decision to Trust Model and Third Space. An explanatory framework is presented covering the micro, meso, and macro levels of CA. Use of a stress and wellbeing lens has allowed insight into the precursors of health in those receiving intensive advice. Using these measures whilst explaining contextual and mechanistic properties, begins to build a complex and real picture of how advice services impact on health.