Support seeking (also called help seeking) is an important metacognitive skill and can be described as ‘a process whereby learners become: i) aware of need for help, ii) decide to seek help, iii) identify potential helper(s), iv) use strategies to elicit help, and v) evaluate the help-seeking episode' (Aleven et al, 2003: 281). Studies that focus on support seeking amongst higher education students tend to focus on the experience of international candidates/ candidates from a particular cultural background and/or the support that PhD candidates seek from their supervisors (for example, Kim, 2007; Curtain, et al., 2013). There are also a growing number of studies reporting first-hand accounts by PhD candidates of their candidature including support seeking. Some of these point to the use of new technologies to seek support outside university boundaries (eg. Grant, 2007; Hermann 2012). There is relatively little research that draws on learner characteristics in conjunction with first-hand accounts, and particularly the relationship between support seeking and learner characteristics. The connection has importance as there is evidence, often untested, that appropriate and timely support seeking is indicative of dealing effectively with the demands of doctoral study and timely completion. This study explores the link between the known learning characteristics of individuals with the help seeking strategies they describe in the five domains outlined above. The aim of this research is to explore support seeking more comprehensively to better articulate how it operate; and to determine if and how different forms of the help-seeking activity contribute to candidate progress early in candidature. The study draws on 40 in depth telephone interviews with early stage (i.e pre-confirmation level) candidates. The participants first completed a survey that drew on i) measures of cognitive functioning, metacognitive self-regulation and management of affect developed by Cantwell et al. (2012), and ii) measures of satisfaction with supervision, department and peer group, and departmental infrastructure provision. On the basis of the survey data we found that candidates fell into one of three clusters. Those whose metacognitive profile suggested strong alignment with the task and strong agency; those with weaker alignment but strong agency, and in the third cluster, weak alignment and weak agency. Patterns of support seeking elicited in the interviews will be reported within the framework of these three groups.