We have previously reported that neonatal (P3) serotonin (5-HT) depletion results in a significant decrease in the number of dendritic spines per 50 microns of dendritic length on dentate granule cells. This effect is specific and permanent. Neither total dendritic length nor the number of dendritic segments is affected by 5-HT depletion. The area dentata contains a dense 5-HT1a receptor population that is present in the at birth. Therefore, 5-HT1a receptors represented a likely candidate for the mediation of the effects of 5-HT on developing granule cells. The present study used the drugs buspirone and NAN-190, which have been shown to be an agonist and antagonist respectively at postsynaptic 5-HT1a receptors in vivo, to test the idea that neurotrophic actions of 5-HT result from 5-HT1a receptor stimulation. Following 5-HT depletion with PCA, pups received daily injections of buspirone (1.0 mg/kg) from P5 to P14. Granule cell morphology was then studied using intracellular filling with Neurobiotin on P14, P21 and P60. Buspirone treatment prevented the loss of dendritic spines previously shown to follow 5-HT depletion with PCA. No other morphological parameters were significantly changed by buspirone treatment. Naive pups received daily injections of NAN-190 from P3 to P14. One group received 1.0 mg/kg while a second group received 3.5 mg/kg. Both doses of NAN-190 resulted in dendritic spine loss comparable to that obtained with neonatal PCA treatment. This loss was permanent suggesting that the first two postnatal weeks may represent a critical period for the action of 5-HT on developing granule cells. Significant, dose-dependent changes in total dendritic length and number of dendritic segments reminiscent of the effects of norepinephrine depletion were also observed in NAN-190-treated rats. We suspect that this change is the result of the action NAN-190 at alpha receptors and is therefore distinct from the specific effect of 5-HT on the number of dendritic spines. The NAN-190 experiment also shows that the loss of dendritic spines is a function of decreased stimulation of 5-HT1a receptors and not the loss of 5-HT terminal membrane.