Background. The goal of this study was to describe the physiological reactions of pregnant women confronted with an experimental psychosocial stressor, and to determine whether the protocol is adequate for stress testing during pregnancy. Methods. Healthy primiparae (n=120) took part in a public speaking and mental arithmetic test. Results. The protocol was found to induce significant increases in saliva cortisol, systolic and diastolic blood pressure and heart rate. The physiological reactions were relatively large compared to those of earlier published studies using physical and psychological stressors during pregnancy, but comparable or slightly lower than those of a group of non-pregnant participants (n=31) tested with a similar protocol. Also, inter-individual variability in reactivity was found, with subjects differing substantially in the magnitude of their reactions to the experimental situation. The cardiovascular responses were correlated to each other and to the overall level of cortisol during the test. Finally, reactivity showed circadian variations with cortisol and systolic blood pressure reactions significantly different (i.e. smaller or even absent) in subjects tested in the morning compared to those tested in the afternoon. Testing in the afternoon hours is recommended. Conclusions. The protocol described in this paper appears to constitute an adequate tool for studying differences in maternal stress reactivity during pregnancy, and as such, can be valuable for studies on maternal prenatal stress.