As an effective method to reduce indoor solar heat gain, louver blinds are increasingly utilized in buildings with high window-to-wall ratios. The blocking of these blinds changes the flow field around the building and then affect the convective heat loss of the facade behind. However, the present correlations utilized to describe the convective heat transfer at the surfaces of the facade and blinds are based on the studies focusing on a smooth plate or a smooth facade of a bluff body, whose accuracy has not been verified. Therefore, in this study, a wind-tunnel experiment is conducted using thermal balance analysis to investigate the convective heat exchange at the blinds and the facade behind. The results show that the current correlations will significantly underestimate the average convective heat transfer coefficient (CHTC) and Nusselt number (Nu) at the blinds while overestimating them at the building facade behind. The results also indicate that the convective heat transfer intensity at the blinds is positively correlated to phi (louver angle) and W (distance between the facade and blinds), whereas that at the facade behind is positively correlated to W, and negatively correlated to phi and theta (temperature difference between the facade and blinds). The natural convection around blinds in summer, which is characterized by the average Ri(B) (Richardson number for the louver blinds), has little effect on the CHTC and Nu at the facade and blinds. This study provides a contribution to clarify the thermal performance of the building envelopes attached with louver blinds.