INTRODUCTION: Dendritic spines were first described by Ramón y Cajal in 1888, and considered by him to be the major sites of axo dendritic apposition and therefore of synaptic input in the CNS. Although a considerable wealth of information has been gathered over the last few decades about the function of spines in the mature nervous system, much less is known about how spines first appear on the otherwise smooth dendritic shafts. DEVELOPMENT: The earliest dendritic appendages, known as filopodia, are long and thin protrusions that occur predominantly during early postnatal development of the mammalian CNS. It is tempting to consider filopodia simply as precursors to spines because at first glance their overall shape is similar to that of mature spines and because their expression during development precedes that of spines. However, the elongated shape of dendritic filopodia (reminiscent of that of axonal filopodia and filopodia in non-neuronal cells) suggests an exploratory function, so that their role may be to contact axons in order to establish early synapses, independently of the eventual formation of spines. CONCLUSIONS: Here we review the literature on dendritic filopodia in an attempt to resolve this issue regarding these two distinct (though potentially overlapping) roles of filopodia in development: spinogenesis vs synaptogenesis. We summarize what is known about the physical characteristics and developmental time course of filopodia expression, as well as the mechanisms of growth and motility of these early dendritic protrusions, both in the intact nervous system and in pathologic settings. Throughout this review we present evidence that supports two hypotheses: that filopodia and spines are two inherently different types of protrusions, and that the role of dendritic filopodia is to capture axons and make early synapses, rather than transform into spines. Finally, we also discuss the potential role of filopodia in the sculpting of the dendritic tree. We also postulate that filopodia have additional important roles in regeneration and repair, in developmental plasticity and in the elaboration of dendritic arbors. These functions may not be limited to a specific developmental period, but probably extend into adulthood. We end by discussing specific experiments that could serve to test these hypotheses.