The core of my research project are Narratives Instances (NIs) of Admission Interviews (AIs) in Public Mental Health carried out in Youth Centres of Mendoza Province, Argentina – within MOdEAS (Modelling of Admission Interviews). The NIs in AIs are oral, spontaneous and co-produced by patients and professionals; therefore, during the initial stage of my research, I redefi ned the categories of narrative genres (Labov, 1972; Labov y Walezky, 1997; Plum, 2004; Martin y Rose, 2008), which were generated on the basis of written narrative texts or induced oral narrative texts, so as to make them applicable to the co-produced and spontaneous orality of NIs in AIs (Salmaso, 2009, 2010, 2012). In this paper, I provide essential evidence to validate my redefinition: there is a systematic relation between the type of constituent and the density of a certain type of process. I classify the constituents into three classes: descriptive constituents – describe characters, time and place –, eventive constituents – present main events and events logically or temporally related to the main events – and evaluative constituents – denote positive or negative evaluation of affect, judgement or appreciation (Martin y White, 2005). I analyse the correlation between these categories of constituents and the different types of processes. The nucleus of my redefinition proposal is the flexibilisation of the categories of NIs through the identification of common and differential constituents between the five categories. I show that the common constituents exhibit the same type of processes in all five ategories of NIs while differential constituents are characterised by having more density of certain types of processes. I herein present the transitivity analysis (Halliday, 1994; Halliday y Matthiessen, 2004; Martin, Matthiessen y Painter, 1997; Matthiessen, 1995) of six Anecdotes out of the twenty five NIs analysed, of eleven speakers, which constitute the corpus of my research – all of them extracted from the corpus of MOdEAS project made up of thirty interviews. The results of the analyses show that eventive constituents have greater density of material processes in independent clauses followed by verbal processes in independent clauses. Descriptive constituents present a greater amount of relational processes in independent clauses and, occasionally, material processes accompanied by circumstances of location and/or time to express habits or to describe the scene where the events take place. The evaluative constituent of Anecdotes, called ‘interpretation of the reaction’, presents special patterns that help characterise it and distinguish it from the other evaluative constituents; it has greater density of verbal processes followed by relational processes. Key words: admission interviews, anecdotes, transitivity, lexical restrictions.