Major depressive disorder is a chronic disabling disease, often triggered and exacerbated by stressors of a social nature. Hippocampal volume reductions have been reported in depressed patients. In support of the neurogenesis theory of depression, in several stress-based animal models of depression, adult hippocampal neurogenesis was reduced and subsequently rescued by parallel antidepressant treatment. Here, we investigated whether repeated social defeat and subsequent individual housing for 3 months induces long-lasting changes in adult hippocampal neurogenesis in rats, and whether these can be normalized by late antidepressant treatment, as would match human depression. Neurogenesis was analysed by stereological quantification of the number of immature doublecortin (DCX)-immunopositive cells, in particular young (class I) and more mature (class II) DCX+ cells, to distinguish differential effects of stress or drug treatment on these subpopulations. Using this social defeat paradigm, the total DCX+ cell number was significantly reduced. This was most profound for older (class II) DCX+ cells with long apical dendrites, whereas younger, class I cells remained unaffected. Treatment with the broad-acting tricyclic antidepressant imipramine, only during the last 3 weeks of the 3-month period after social defeat, completely restored the reduction in neurogenesis by increasing both class I and II DCX+ cell populations. We conclude that despite the lack of elevated corticosterone plasma levels, neurogenesis is affected in a lasting manner by a decline in a distinct neuronal population of more mature newborn cells. Thus, the neurogenic deficit induced by this social defeat paradigm is long-lasting, but can still be normalized by late imipramine treatment.