Background: The electrophysiological and behavioral studies have shown that saffron extract improves memory and learning skills in ethanol-induced learning behavior impairments. Objective: The aim of the present study was to evaluate the possible effects of Crocus sativus aqueous extract stigma and its constituents, safranal and crocin, on intact memory and scopolamine-induced learning deficits in rats performing the Morris water maze task. Method: The extract and its constituents were injected intraperitoneally to rats one h before training for five consecutive days. During the training period, four trials were carried out each day. On the fifth day of the experiment, the locomotor activity was assessed using open field test. Results: Generally, the aqueous extract (0.0025-0.56 g/kg) of stigma did not decrease the latency time to find the platform. Only a dose of 0.005 g/kg reduced the latency time during five days. Crocin (50 and 200 mg/kg) also did not change the latency time. Safranal (0.2-0.75 ml/kg) increased the time to find the platform. The effects of the extract, crocin and safranal on hyoscine-induced learning deficits were also evaluated (1 and 500 mg/kg hyoscine). The stigma extract and crocin reduced the deficit effect of hyoscine (in low and high dose) on memory. Safranal (except a dose of 0.2 ml/kg) did not decrease the effect of hyoscine on memory. In the open field activity test, the aqueous extract, crocin and safranal reduced total activity. The co-administration of these agents with hyocine significantly compensated the reduction of locomotor activity in hyosine group alone. Conclusion: On the basis of these results, the Crocus sativus stigma aqueous extract, crocin and safranal did not have effect on intact memory but the extract and crocin inhibited the hyoscine impaired acquisition/performance activity. These agents can be potentially introduced as new drugs in the treatment of cognitive dysfunctions such as Alzheimer's disease.