Employability can be constituted by a combination of individuals' characteristics and the state of the labour market (Brown et al., 2003). Thus, employability includes individual factors, personal circumstances and external factors (McQuaid & Lindsay, 2005). It can also be defined solely from an individual point of view emphasising the achievements, understandings and personal attributes that make individuals more likely to gain employment and be successful in their chosen professions (Knight & Yorke, 2002, 2003). Most research on the topic has explored students' employability and transition from university to work by concentrating on the relationship between the output of academic education and the demands of working life, as well as employers' expectations (e.g. Kavanagh & Drennan, 2008). Moreover, at the institutional level, the employability of graduates is often measured by graduate employment rates (e.g. Harvey, 2001). However, to find employment after their university studies, graduates must be able to identify their competences and be confident that they will succeed in working life. It is therefore important to explore the transition from university to work from the graduates' perspective (Johnston, 2003). Knight and Yorke (2002) have suggested a model for excellent employability. The model consists of four interrelated components: Understanding, Skills, Efficacy beliefs and Metacognition (i.e. the USEM model). In line with the USEM model, Bennett, Dunne and Carré (1999) developed a model of course provision to develop students' generic skills and employability. It consists of both disciplinary content and skills and generic skills, but also workplace experience and workplace awareness. In this study, we use the concept ‘academic competences’; these include academic skills such as critical thinking, collaboration and communication skills, and problem-solving skills (e.g. Strijbos et al., 2015). Such skills can be developed and utilised in both university and work contexts (Bennett et al., 1999; Greenbank et al., 2009). Academic competence can therefore be defined as a multidimensional construct 239of skills, attitudes and behaviours, including academic skills, study skills, academic motivation, interpersonal skills and academic self-conceptions (DiPerna & Elliott, 1999). In line with Knight and Yorke (2002) and Bennett et al. (1999), the aim of the present study is to capture the individual-level variation in graduates' evaluations of their academic competences, their confidence in becoming successful in working life, and the usefulness of work experience to their studies during the period of transition to working life, in order to gain deeper insight into the employability of graduates. Next, the elements of the USEM model and the model of Bennett et al. (1999) are discussed in more detail, starting from ‘understanding and skills’.