Social support is strongly associated with resilience and positive outcomes in children who have experienced risk and adversity, including children and young people in care. However, research suggests that children in care are often disadvantaged in their ability to benefit from social support due to disrupted relationships and multiple placements. Whilst a number of measures have been developed to assess the support networks of children, few are able to adequately assess and describe the complex network of relationships that exist for young people in care.The aim of this exploratory study was to investigate the use of a Dependency Grid to assess social support in young people who are looked after. The Dependency Grid is an interview technique derived from George Kelly's (1955) Theory of Personal Constructs and was administered to ten individuals aged 11-17 years who were either in foster care or residential care. Qualitative and quantitative methods of data analysis were used to explore its utility to assess social support with each case. The appropriateness and ease of completing the Dependency Grid was evaluated through participant feedback interviews. The views and opinions of the young people's social workers and a reference group consisting of social care practitioners were also obtained.The Dependency Grid was found to be a useful tool for examining the support networks of young people in care, with many advantages over existing measures. The Dependency Grid revealed similarities and differences in the way participants distributed their dependencies. Most of the participants had dispersed dependency grids and carers and friends were the most significant providers of support. Levels of social support did not vary with the increasing size of the support network (r = 0.198) and the felt 'closeness' of relationships, as measured using the Four Field Map, was poorly related to the allocation of dependencies (r = -0.243). Some evidence was found in relation to the validity of the Dependency Grid for individual cases using qualitative methods. The strengths and limitations of the Dependency Grid are discussed and the methodological challenges with the study and opportunities for further research are considered.