This study was designed to examine the sequential changes in the developing granuloprival cerebellar culture. In this model of anomalous cerebellar development, organotypic cultures derived from newborn Swiss-Webster mice were exposed to the DNA synthesis inhibitor, cytosine arabinoside, at explantation and were fixed for electron microscopic examination on successive days in vitro. Similar developmental stages were compared in control explants. Granule cell destruction began early, and was widespread by 2 days in vitro, when oligodendrocyte destruction also began in treated cultures. A few granule cells survived, but no recognizable oligodendrocytes remained by 7 days in vitro, at a time when myelin was initially evident in control explants. Purkinje cell recurrent axon collateral sprouting began at 3 days in vitro in cultures exposed to cytosine arabinoside, and the sprouted terminals initially synapsed with Purkinje cell somata, somatic spines and dendritic shafts. Synapses with Purkinje cell dendritic spines developed later, at approximately the same time as parallel fiber-Purkinje cell dendritic spine synapses formed in control cultures. Astrocytic ensheathment of control Purkinje cells was well underway by 6 days in vitro and Purkinje cell somata were relatively rounded and almost completely ensheathed by 9 days in vitro. Glial ensheathment did not occur in cytosine arabinoside treated cultures, and Purkinje cell somata were scalloped at 7 days in vitro by excess impinging recurrent axon collateral terminals, and never developed the smooth contours characteristic of control Purkinje cells. Purkinje cell somatic spines persisted in treated explants, and reduction of excess extracellular space was delayed until 12 days in vitro, when most of the developmental changes had been completed. The earlier development of synapses by excess recurrent axon collateral terminals with Purkinje cell somata, somatic spines and dendritic shafts, followed by the later development of heterotypical synapses with dendritic spines, in parallel with synapse formation by normal presynaptic elements, suggests that the sequence of development of synapses is a function of the maturational state of the postsynaptic components.