Chronic cocaine exposure influences the density and morphology of dendritic spines on medium spiny neurons (MSNs) in the nucleus accumbens (NAc), a critical brain region for cocaine craving. However, the relationship between spine plasticity and craving remains unclear. To study this relationship, we trained rats to self-administer cocaine using an extended-access regimen (6 h per day, 10 days); controls self-administered saline. Previously, a time-dependent intensification (incubation) of cue-induced cocaine craving has been demonstrated after withdrawal from this regimen; furthermore, Ca2+-permeable AMPA receptors (CP-AMPARs) increase in the NAc core after ~1 month of withdrawal and thereafter mediate the expression of incubated craving. Although neither craving nor CP-AMPAR levels were measured in the present study, we killed rats at four withdrawal day (WD) time-points (WD14, WD25, WD36, or WD60) selected to span the rising phase of incubation and the transition from low to high CP-AMPAR levels. MSNs were iontophoretically filled with Lucifer yellow and spines were analyzed with NeuronStudio software. Compared with saline controls, cocaine rats showed no changes in spine density or morphology in the NAc core on WD14 or WD25. On WD36, approximately the withdrawal time when stable elevation of CP-AMPAR levels is detected, the cocaine group exhibited increased density of thin spines in the NAc core. By WD60, however, this effect had reversed: the density of thin spines was lower in cocaine rats compared with saline rats. In contrast, craving and CP-AMPAR levels remain high on WD60. We also assessed spine density on WD36 in the dorsolateral striatum, a region that is not implicated in incubation of cocaine craving and does not undergo CP-AMPAR plasticity. Here, the cocaine group exhibited a small leftward shift in the distribution of spine densities plotted as a cumulative distribution, opposite to the effect found in the NAc core. Overall, our results demonstrate changes in NAc core spines over 2 months of withdrawal but no simple relationship between the time dependency of these spine changes and the previously demonstrated time course of incubation of cocaine craving. However, they raise the possibility that CP-AMPAR accumulation in the NAc core occurs in a population of thin spines that emerges after ~1 month of withdrawal.