The United Nations have identified climate change as the greatest threat to human life. As current research shows, urban areas are more vulnerable to climate change than rural areas. Numerous people are affected by climate change in their daily life, health and well-being. The need to react is undisputed and has led to numerous guidelines and directives for urban climate adaptation. Plants are commonly mentioned and recommended as one key to urban climate adaptation. Due to shading of open space and building surfaces, as well as evapotranspiration, plants reduce the energy load on the urban fabric and increase thermal comfort and climate resilience amongst many other ecosystem services. Plants, therefore, are described as green infrastructure (GI), because of the beneficial effects they provide. Extensive green roofs are often discussed regarding their impact on thermal comfort for pedestrians and physical properties of buildings. By means of Stadslab2050 project Elief Playhouse in Antwerp, Belgium, a single-story building in the courtyard of a perimeter block, the effects of different extensive green roof designs (A and B) on the microclimate, human comfort at ground and roof level, as well as building physics are analyzed and compared to the actual roofing (bitumen membrane) as the Status Quo variant. For the analyses and evaluation of the different designs the innovative Green Performance Assessment System (GREENPASS®) method has been chosen. The planning tool combines spatial and volumetric analyses with complex 3D microclimate simulations to calculate key performance indicators such as thermal comfort score, thermal storage score, thermal load score, run-off and carbon sequestration. Complementary maps and graphs are compiled. Overall, the chosen method allows to understand, compare and optimize project designs and performance. The results for the Elief Playhouse show that the implementation of green roofs serves a slight contribution to the urban energy balance but a huge impact on the building and humans. Variant B with entire greening performs better in all considered indicators, than the less greened design Variant A and the actual Status Quo. Variant B will probably bring a greater cost/benefit than Variant A and is thus recommended.