In vivo membrane fusion primarily occurs between highly curved vesicles and planar membranes. A better understanding of fusion entails an accurate in vitro reproduction of the process. To date, supported bilayers have been commonly used to mimic the planar membranes. Soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins that induce membrane fusion usually have limited fluidity when embedded in supported bilayers. This alters the kinetics and prevents correct reconstitution of the overall fusion process. Also, observing content release across the membrane is hindered by the lack of a second aqueous compartment. Recently, a step toward resolving these issues was achieved by using membranes spread on holey substrates. The mobility of proteins was preserved but vesicles were prone to bind to the substrate when reaching the edge of the hole, preventing the observation of many fusion events over the suspended membrane. Building on this recent advance, we designed a method for the formation of pore-spanning lipid bilayers containing t-SNARE proteins on Si/SiO2 holey chips, allowing the observation of many individual vesicle fusion events by both lipid mixing and content release. With this setup, proteins embedded in the suspended membrane bounced back when they reached the edge of the hole which ensured vesicles did not bind to the substrate. We observed SNARE-dependent membrane fusion with the freestanding bilayer of about 500 vesicles. The time between vesicle docking and fusion is ∼1 s. We also present a new multimodal open-source software, Fusion Analyzer Software, which is required for fast data analysis.