Linguistic and cultural diversity - amongst other things as a result of immigration - is a common reality in German classrooms, particularly in urban areas. That raises the question, to what extent and in what way the circumstance that pupils with different linguistic and cultural backgrounds are taught together in the same classes affects the conditions of learning and teaching. Up to now, little is known about the impact of the linguistic and cultural background on pupils' learning of mathematics. The empirical study "Learning Mathematics in the Context of Linguistic and Cultural Diversity" is concerned with this problem. The project draws on data from approximately 120 pupils aged 12 to 14 in six different seventh grade maths classes in Hamburg. Two of the participating schools have a high proportion of Turkish immigrant pupils, another two a significant number of Russian immigrants, while the last two schools predominantly consist of pupils from German families without immigrant background. The overall study is subdivided into three parts with different focuses: the aim of the first part is to explore the cultural and social background of the pupils' families, part two ascertains the pupils' linguistic proficiency, and the third part finally attends to the pupils' "mathematical habitus" - that is, their school-mathematics related practices and representations. The latter part is outlined in the following. It is quality-based and the collection and analysis of data and the development of theory follows methods of the Grounded Theory. Firstly, the study is based on current knowledge about mathematical beliefs from the field of mathematics education. Additionally, Pierre Bourdieu's concept of "habitus" is used. It is employed as a model for describing and analysing the impact of pupils' linguistic and cultural backgrounds on their mathematics learning. Empirical data were collected from classroom observations, mathematical case studies on problem solving activities with selected pupils, and also surveys with questionnaires and interviews with pupils as well as with parents. First findings of the investigations indicate remarkable differences in the manner the participating pupils perceive and solve mathematical problems, and there is evidence that these differences are related to the pupils' linguistic and cultural background. In this paper - intended as a kind of work- in-progress report - an outline of the theoretical background of the study, as well as an illustration and interpretation of preliminary findings will be presented.