The glutamatergic synapses of developing neuromuscular junctions (NMJ) of Drosophila larvae are readily accessible, morphologically simple, and physiologically well-characterized. They therefore have a long and highly successful tradition as a model system for the discovery of genetic and molecular mechanisms of target recognition, synaptogenesis, NMJ development, and synaptic plasticity. However, since the development and the activity-dependent refinement of NMJs are concurrent processes, they cannot easily be separated by the widely applied genetic manipulations that mostly have chronic effects. Recent studies have therefore begun systematically to incorporate larval foraging behavior into the physiological and genetic analysis of NMJ function in order to analyze potential experience-dependent changes of glutamatergic transmission. These studies have revealed that recent crawling experience is a potent modulator of glutamatergic transmission at NMJs, because high crawling activities result after an initial lag-phase in several subsequent phases of experience-dependent synaptic potentiation. Depending on the time window of occurrence, four distinct phases of experience-dependent potentiation have been defined. These phases of potentiation can be followed from their initial induction (phase-I) up to the morphological consolidation (phase-III/IV) of previously established functional changes (phase-II). This therefore establishes, for the first time, a temporal hierarchy of mechanisms involved in the use-dependent modification of glutamatergic synapses.