Working condition assessment is required by Dutch law to obtain approval for occupying building. Nowadays this is a manual procedure that is executed after the building has been realized. Experts are needed to interpret the legislation regulations and test these against what they observe. If this procedure is integrated in the design process, then the risk of costly building changes can be reduced. In this article we describe a system that supports automated testing before building construction using fuzzy logic for reasoning on the working condition regulations. The intrinsic character of the regulations often makes it very difficult to use classical rule-based systems. Instead of stating hard and unambiguous demands, the law contains many rules that are open for discussion, that allow alternate solutions or that are simply put rather vague. In this article we demonstrate that fuzzy logic provides a methodology to formalize such regulations. For testing working conditions, input data are needed about the building geometry and material use, building physics (e.g. lighting) and space utilisation. We describe which data are required and how they can be generated using state-of-the-art research and technology. In a case study, the process of implementing typical regulations shows the implications of implementation automated working conditions checking in practice.