The aim was to evaluate the cardiovascular and subjective stress response to a combined physical and mental workload, and the effect of rest. Twelve females who had no prior experience of laboratory experiments participated in the study. Computer-work-related mental stressors were either added to or removed from a standardized computer work session in the laboratory. Beat-to-beat blood pressure and electrocardiogram (ECG) were recorded continuously during the experiment. The participants reported subjective experiences of stress in six categories using an 11-point scale before and at the end of the work. Heart rate variability (HRV) variables were calculated from the ECG recordings, and a reduction in the high-frequency component of HRV and an increase in the low- to high-frequency ratio were observed in the stress situation compared to the control session. No changes were seen in the low-frequency component of HRV. The stressors induced an increase in blood pressure compared to baseline that persisted, and for the diastolic pressure it even increased in the subsequent control session. No differences were observed for subjective experience of stress with the exception of a time trend in the exhaustion scale, i.e. a progression in reported exhaustion with time. The results-and the dissociation between HRV and blood pressure variables-indicate that HRV is a more sensitive and selective measure of mental stress. It could be speculated that heart rate-derived variables reflect a central pathway in cardiovascular control mechanisms ("central command"), while the blood pressure response is more influenced by local conditions in the working muscles that partly mask the effect of changes in mental workloads. In the rest period after each work session, HRV and blood pressure variables were partly normalized as expected. However, an 8-min period of rest was insufficient to restore blood pressure to resting values.