This article draws on Gee's notions of Discourse and specifically Discourse Models to explore Engineering Problem Solving1 and the different ways in which it can be understood in an engineering context. After Gee we attempt to identify aspects of doing, being and valuing that underpin people's Problem Solving Discourse Models. Interviews with three engineering lecturers reveal that they draw extensively on two different Discourse Models of Engineering Problem Solving. The more highly valued Model (Integrated Design Model) reflects engineering practice, is located in engineering design and dependant on judgement. The other is located in the classroom and involves the algorithmic resolution of mathematical models, (Knowledge Construction Model). These Discourse Models form a backdrop to interviews with three students entering an engineering degree programme for the first time. The three students each draw different Discourse Models of Problem Solving, and display characteristics (such as the level of confidence) that align more or less with Engineering Problem Solving, sometimes obscuring their understanding. The implications of these findings in terms of an introductory engineering course are discussed. These include recognising the potential diversity of Problem Solving Discourse Models our students bring to tertiary education, as well as the difficulty of introducing a legitimate design project requiring the level of judgement needed to interpret open-ended, ill-defined problems and then integrate multiple quantitative models with multidisciplinary qualitative judgements in a rigorous manner.