The Oxford Admissions Study aims to contribute to the discourse regarding admissions to universities in general and admissions to the University of Oxford in particular. The project was set up to examine whether any social background factors have a systematic effect on the likelihood of being admitted to Oxford University once controls for 'academic merit' are applied. Methods and Data During the admissions round to Oxford University for entry in 2003, 1928 candidates for admissions at selected Oxford colleges participated in the Oxford Admissions Study. As part of the project, candidates filled in a survey that included motivational measures as well as measures of social and cultural capital, expectations and aspirations, revision strategy and parental socio- economic and educational background. In addition, the Alice Heim reasoning test was administered to assess whether it added any information to the information on GCSE and predicted A-levels. It was hoped that this measure might assess 'potential' rather than achievement and that it thus might be relatively uncorrelated to school attainment. About half of the survey participants also participated in a deep and surface learning exercise which involved the analysis of a scientific text. For the part of the project that I hope to present at the ECER, the method of analysis will be regression models predicting the odds of being offered a place at the University of Oxford. Analysis and Findings In line with previous research, the analysis finds gross differences in the likelihood of being accepted by ethnicity and, to a lesser extent, by social class. Previous research in this area had to rely on actually achieved A-level grades reported by UCAS and found that these differences persist once these measures of merit are added to the model (net difference). Differences in the likelihood of acceptance without adding the 'meritocratic' controls are not necessarily worrying for the University of Oxford though they feedback into the greater debate of the link between social background and educational attainment. In contrast, net differences in the likelihood of gaining an offer after controlling for measures of achievement should not exist if 'perfect meritocracy' (as understood by school attainment) was practised in the admissions process. Our research shows that the most significant predictor of gaining an offer is GCSE attainment. It works to candidates disadvantage to have lower predictions than straight As for their A-levels. We find that the strength of the link between certain social background charachteristics and admissions decisions is indeed decreased once we control for 'merit'. However, we also find that some disadvantages persist even when academic merit is controlled for. Future outlook The project aims to follow all survey participants, those admitted to Oxford as well as those who subsequently attended other universities, throughout their degree. It is our aim to collect students' degree results and to link those back to measures of aspiration and motivation collected at the beginning of the project. We therefore hope that this project can also shed some light on the persistence of the 'gender gap' in finals at the University of Oxford.