This study examined the modulatory effects of tonic and phasic variations in blood pressure on working memory. Newly-diagnosed unmedicated hypertensives (n=24) and normotensives (n=46) performed a working memory task under two memory loads (0-back, 2-back) while 128-channel EEG activity was recorded. Targets and non-targets were presented with equal probability at six delays after the ECG R-wave (100 to 600 ms). Difference ERP waveforms (stimulus-locked minus previous-beat-locked) were computed to reduce cardiac and cardiovascular contributions to the ERP signals. Permutation testing revealed that the hypertensives showed larger mid- and late-latency ERP positivity than the normotensives despite no overall behavioral performance differences. Cardiac cycle time analyses indicated that working memory performance was less accurate for stimuli presented at 300 ms and 500 ms after the R-wave for both groups but only under higher memory load. Both groups showed cardiac cycle effects involving mid- and late-latency ERP for stimuli presented at 300 ms and 500 ms after the R-wave. However, fewer effects emerged in the hypertensives. Our study reveals that the phase of the cardiac cycle modulates working memory processing. This modulation may be attributable to baroreceptor-related cortical inhibition. The reduced cardiac cycle modulation in hypertensives fits with their decreased baroreflex sensitivity. In the absence of noticeable cognitive decline, ERP indices and cardiac cycle effects can identify early manifestations of functional deficits in hypertension.