The work presented in this thesis concerns the development of a viscous-inviscid interaction model for prediction of viscous flow properties in aeronautical applications. The inviscid model is based on an existing potential flow model and the thesis thus focuses on the development of a viscous model based on the three-dimensional integral boundary layer equations. The model is to be applicable to complex geometries with unstructured meshes and this requirement, in combination with the fairly complex character of the integral boundary layer equations, sets high standards for the discretization. In order to arrive at a stable and well-conditioned scheme a number of topics related to the formulation and discretization of the integral model are analyzed and discussed. These include the challenge of finding a discretization which ensures flow conservation on curved surfaces and provides a stable and well-conditioned discretization for mixed-hyperbolic systems of conservation laws. Different coupling strategies for the viscous and inviscid models are discussed and analyzed and so are the singularities in the integral boundary layer equations in separated flow regions. The predictions of the resulting viscous-inviscid coupling scheme are validated by comparison to experimental measurements as well as to predictions from other numerical models. The currently developed coupled model is found to provide reasonably accurate predictions of viscous flow properties in laminar as well as turbulent flow regions while being stable and convergent in separated flow regions.