Not all antibodies against SARS-CoV-2 inhibit viral entry, and hence, infection. Neutralizing antibodies are more likely to reflect real immunity; however, certain tests investigate protein/protein interaction rather than the fusion event. Viral and pseudoviral entry assays detect functionally active antibodies but are limited by biosafety and standardization issues. We have developed a Spike/ACE2-dependent fusion assay, based on a split luciferase. Hela cells stably transduced with Spike and a large fragment of luciferase were co-cultured with Hela cells transduced with ACE2 and the complementary small fragment of luciferase. Cell fusion occurred rapidly allowing the measurement of luminescence. Light emission was abolished in the absence of Spike and reduced in the presence of proteases. Sera from COVID-19-negative, non-vaccinated individuals or from patients at the moment of first symptoms did not lead to a significant reduction of fusion. Sera from COVID-19-positive patients as well as from vaccinated individuals reduced the fusion. This assay was more correlated to pseudotyped-based entry assay rather than serology or competitive ELISA. In conclusion, we report a new method measuring fusion-inhibitory antibodies in serum, combining the advantage of a complete Spike/ACE2 interaction active on entry with a high degree of standardization, easily allowing automation in a standard bio-safety environment.