The strength of synaptic connections can undergo long-lasting changes, and such long-term plasticity is thought to underlie higher brain functions such as learning and memory. De novo synthesis of proteins is required for such plastic changes. This model is now supported by several lines of experimental data. Components of translational machinery have been identified in dendrites, including ribosomes, translation-al factors, numerous RNAs, and components of posttranslational secretory pathways. Various RNAs have been shown to be actively and rapidly transported to dendrites. Dendritic RNAs typically contain transport-specifying elements (dendritic targeting elements). Such dendritic targeting elements associate with trans-acting factors to form transport-competent ribonucleoprotein particles. It is assumed that molecular motors mediate transport of such particles along dendritic cytoskeletal elements. Once an mRNA has arrived at its dendritic destination site, appropriate spatiotemporal control of its translation, for example, in response to transsynaptic activity, becomes vital. Such local translational control, recent evidence indicates, is implemented at different levels and through various pathways. In the default state, translation is assumed to be repressed, and several mechanisms, some including small untranslated RNAs, have been proposed to contribute to such repression. Translational control at the synapse thus provides a molecular basis for the long-term, input-specific modulation of synaptic strength.