Female labour force participation and fertility rates have been very low in Germany, particularly among women with a university degree. Long career breaks and their high opportunity costs may explain both facts. We intend to develop a life-cycle model that can account for these facts and to use it to evaluate the effectiveness of various family policy measures. Educational choices are assumed to be fully forward-looking. In particular, they take into account desired fertility, time to degree and the opportunity costs of career breaks. These costs are specific to occupations and educational achievement because of varying returns to experience. In the empirical implementation, we put particular emphasis on preference heterogeneity and on a detailed representation of the German policy environment. We will estimate the model using a novel administrative dataset that allows a number of key improvements over the current state of the art. After validating the model predictions out-of-sample using several reforms from the past 20 years, we will first evaluate the components of a 2007 reform to parental leave and benefits regulations and their long-term impacts on female labour supply and on fertility rates. Subsequently, we will introduce family policy measures from France and the UK into the tax-transfer system of the model, making full use of its potential to evaluate counterfactual reforms.