This thesis investigates the role of presynaptic activity during embryonic and larval development at the Drosophila neuromuscular junction (NMJ). Electrical activity has been implicated in various developmental steps including growth cone motility, elimination of off-targets synaptic contacts, and synaptic maturation. However, the molecular mechanisms governing these activity-dependent processes are poorly understood. The Drosophila NMJ provides a powerful, stereotyped in vivo model that can help begin to elucidate some of these mechanisms, as reviewed in Chapter I. Chapter II describes the characterization of Shaker Dominant Negative (SDN) to spatially control excitability in the pre- or postsynaptic cell. Through electrophysiological recordings, we demonstrate that SDN enhances excitability by delaying repolarization. Evoked postsynaptic potentials appear broadened when expressing SDN pre- or postsynaptically, mostly likely due to a decrease in the fast acting IA current. In addition, we show that activity-dependent changes in the size of NMJ occur only when SDN is expressed presynaptically. Chapter III investigates the role of presynaptic activity in semaphorin-induced synaptic pruning. During embryonic development, the motoneuron growth cones encounter various chemotropic cues that not only provide guidance but also participate in the stabilization and destabilization of the synapse. Reduction of neuronal activity using mutations affecting sodium channels produces ectopic neuromuscular contacts throughout the bodywall and a similar phenotype is observed in semaphorin-2a mutants. We show that motoneuronal activity and the semaphorin chemorepellent interact to regulate the destabilization and withdrawal of ectopic synapses. A link between the function of voltage-gated calcium channels and the chemorepellent suggests that calcium functions downstream of activity. Further, we implicate a novel function for the calcium/calmodulin kinase II (CaMKII) in the regulation of synaptic refinement. Finally, in Chapter IV, we discuss some of the pertinent issues that arise from our study on refinement, and we propose key experiments to help further elucidate how activity is translated into a molecular signaling cascade.