Brief pulse electroconvulsive therapy (BP ECT; pulse width 0.5-1.5 msec) is a very effective treatment for severe depression but is associated with cognitive side-effects. It has been proposed that ultrabrief pulse (UBP; pulse width 0.25-0.30 msec) ECT may be as effective as BP ECT but have less cognitive effects because it is a more physiological form of neuronal stimulation. To investigate this further, we treated normal rats with a 10 session course of either BP (0.5 msec), UBP (0.3 msec), or sham electroconvulsive stimulation (ECS) and measured antidepressant-related changes in dentate gyrus cell proliferation and hippocampal BDNF protein levels as well as hippocampal-dependant spatial reference memory using the water plus maze and immobility time on the forced swim test. Both BP and UBP ECS induced very similar types of motor seizures. However, BP ECS but not UBP ECS treatment led to a significant, near 3-fold, increase in cell proliferation (p=0.026) and BDNF levels (p=0.01). In the forced swim test, only BP ECS treated animals had a significantly lower immobility time (p=0.046). There was a trend for similarly reduced hippocampal-dependent memory function in both BP and UBP groups but overall there was not a significant difference between treatment and control animals when tested 10 days after completing allocated treatment. These findings show that, even though both forms of ECS elicited similar motor seizures, UBP ECS was less efficient than BP ECS in inducing antidepressant-related molecular, cellular and behavioural changes.