A few experimental studies reported a close association between the frequency of use of mental state terms and theory of mind performance (Hughes et al., 1998; Hughes et al., 2007), while the relationship between mental state talk and mentalizing abilities in spontaneous contexts is less explored. Furthermore, the majority of studies in this field deal with infancy and childhood, and less attention has been devoted to more advanced ages. This research aims at testing the association between pre-adolescents’ mentalizing behaviour, and the ability to use mental state talk in explaining people’s behaviour. The participants were 42 preadolescents (aged 12 years) from middle SES high schools. The Child Attachment Interview (Shmueli-Goetz et al., 2008) was used to explore mental state talk. A vignette-based semi-structured interview (O’Connor et al., 1999) was administered in order to evaluate mentalizing abilities. Three levels in mentalizing responses were discerned, in conformity with the authors’ classification. Verbal intelligence was assessed using the verbal subscale of WISC-III. Results indicated a high correlation between mentalization abilities and mental state talk, either according to quantity or to variety. In addition, there was no association between mental state talk and verbal abilities, nor between mental state talk and gender.