In prior studies, we described the differential organization of corticostriatal and thalamostriatal inputs to the spines of direct pathway (dSPNs) and indirect pathway striatal projection neurons (iSPNs) of the matrix compartment. In the present electron microscopic (EM) analysis, we have refined understanding of the relative amounts of cortical axospinous vs. axodendritic input to the two types of SPNs. Of note, we found that individual dSPNs receive about twice as many axospinous synaptic terminals from IT-type (intratelencephalically projecting) cortical neurons as they do from PT-type (pyramidal tract projecting) cortical neurons. We also found that PT-type axospinous synaptic terminals were about 1.5 times as common on individual iSPNs as IT-type axospinous synaptic terminals. Overall, a higher percentage of IT-type terminals contacted dSPN than iSPN spines, while a higher percentage of PT-type terminals contacted iSPN than dSPN spines. Notably, IT-type axospinous synaptic terminals were significantly larger on iSPN spines than on dSPN spines. By contrast to axospinous input, the axodendritic PT-type input to dSPNs was more substantial than that to iSPNs, and the axodendritic IT-type input appeared to be meager and comparable for both SPN types. The prominent axodendritic PT-type input to dSPNs may accentuate their PT-type responsiveness, and the large size of axospinous IT-type terminals on iSPNs may accentuate their IT-type responsiveness. Using transneuronal labeling with rabies virus to selectively label the cortical neurons with direct input to the dSPNs projecting to the substantia nigra pars reticulata, we found that the input predominantly arose from neurons in the upper layers of motor cortices, in which IT-type perikarya predominate. The differential cortical input to SPNs is likely to play key roles in motor control and motor learning.