The cytoplasmic H3 helical domain of syntaxin is implicated in numerous protein-protein interactions required for the assembly and stability of the SNARE complex mediating vesicular fusion at the synapse. Two specific hydrophobic residues (Ala-240, Val-244) in H3 layers 4 and 5 of mammalian syntaxin1A have been suggested to be involved in SNARE complex stability and required for the inhibitory effects of syntaxin on N-type calcium channels. We have generated the equivalent double point mutations in Drosophila syntaxin1A (A243V, V247A; syx(4) mutant) to examine their significance in synaptic transmission in vivo. The syx(4) mutant animals are embryonic lethal and display severely impaired neuronal secretion, although non-neuronal secretion appears normal. Synaptic transmission is nearly abolished, with residual transmission delayed, highly variable, and nonsynchronous, strongly reminiscent of transmission in null synaptotagmin I mutants. However, the syx(4) mutants show no alterations in synaptic protein levels in vivo or syntaxin partner binding interactions in vitro. Rather, syx(4) mutant animals have severely impaired hypertonic saline response in vivo, an assay indicating loss of fusion-competent synaptic vesicles, and in vitro SNARE complexes containing Syx(4) protein have significantly compromised stability. These data suggest that the same residues required for syntaxin-mediated calcium channel inhibition are required for the generation of fusion-competent vesicles in a neuronal-specific mechanism acting at synapses.