Neuronal function mainly depends on the number of established synapses and the release probability per synapse. It has been long debated whether these synaptic features are regulated by a single mechanism or result from independent processes that together sustain neuron homeostasis. The Drosophila Ca2+-sensor Frequenin – duplicated to Frq1 and Frq2, but a single protein in mammals, neuronal calcium sensor 1 (NCS-1) – has been found to be involved in both synaptic processes, but its exact biological roles remain unclear. On page 4246, Alberto Ferrús and colleagues first perform a yeast two-hybrid screen to isolate the cytosolic guanine-exchange factor (GEF) Ric8a as a new Frq-interacting protein. Interestingly, they find that Ric8a only interacts with Frq2, not with the virtually identical Frq1, and determine R94 and T138 as key amino acid residues responsible for this difference. This interaction is conserved in human Ric8a and NCS-1, which could be important as defects in NCS-1 have been linked to mental retardation and autism. Mechanistically, Ric8a regulates both synapse number and neurotransmitter release, whereby the former effect is mediated through binding to Gαs – a Gα subtype – and negatively regulated by Frq2 to control synapse number, and the latter is independent of Frq2. Thus, the data presented here point to a common pathway, Frq2–Ric8a–Gαs, in controlling both synaptic functions, which then diverges downstream to regulate neuron homeostasis.