Camillo Golgi and Santiago Ramón y Cajal famously argued over whether the nervous system was one big syncytium or instead a network of independent units. While it is clear that neurons are separate units as Cajal suggested, communication between neurons can occur directly, without chemical intermediary, at electrical synapses, akin to what Golgi envisioned. Electrical synapses are composed of gap junction (GJ) channels between neurons allowing for direct ionic and metabolic communication. Neuronal GJs are often thought to be simple, symmetric structures, with the same protein constituents on either side of the synapse. During GJ formation, both the pre- and postsynaptic neurons contribute hemichannels composed of hexamers of Connexin (Cx) protein to form functional channels between neurons. Despite this apparent simplicity, work from cell culture has found that GJs can contain multiple different Cx proteins and this can affect the functional properties of communication. In a forward genetic screen in zebrafish, which looked for mutations affecting electrical synapse formation in the Mauthner neural circuit, we identified the Dis3 mutation that disrupts a homologue of mammalian connexin36 (cx36). cx36 is thought of as the main neuronal GJ gene due to its broad neuronal expression and extensive contributions to electrical synapses throughout the brain,. Through genome gazing we found that zebrafish have four orthologous cx36-like genes, cx34a (Dis3), cx34b, cx35a, and cx35b. To identify which Cxs were required we developed a CRISPR-based in vivo screen and found that only cx34a and cx35a were necessary for M electrical synapse formation. Using cell transplantations we found that cx34a is required exclusively postsynaptically, while cx35a is required exclusively presynaptically, for synaptogenesis. Additionally, the CRISPR screen identified a scaffolding molecule, tjp1b, that is necessary for electrical synaptogenesis and preliminary evidence suggests tjp1b is required exclusively postsynaptically. We conclude that vertebrate electrical synapses can be molecularly asymmetric at the level of the GJ proteins themselves. In addition, our preliminary evidence suggests that such asymmetries can extend beyond the channel to the larger proteome of the electrical synapse. Such asymmetries at the GJ and beyond suggest a molecular mechanism for generating functional asymmetry across the electrical synapse. Moreover, this work suggests that we are just starting to scratch the surface of electrical synapse complexity and future work will continue to reveal the richness of these structures at the molecular and functional level. Also, it hopefully allows Golgi to rest a bit easier.