Background: Twenty-four-hour blood pressure variability (BPV) is independently related to cardiovascular outcomes, but limited and conflicting evidence is available on the relative prognostic importance of systolic and diastolic BPV. The aim of this study was to verify the hypothesis that the association of systolic and diastolic blood pressure variability over 24 h with cardiovascular mortality in untreated subjects is affected by age. Design and methods: The study included 9154 untreated individuals assessed for hypertension between 1982 and 2002 in the frame of the Dublin Outcome Study, in which 24 h ambulatory blood pressure monitoring was obtained (age 54.1 ± 14.3 years, 47% males). The association of short-term systolic and diastolic blood pressure variability with cardiovascular and all-cause mortality in the entire sample and separately in younger and older age subgroups was assessed over a median follow-up period of 6.3 years. Results: Diastolic BPV was directly and independently related to cardiovascular mortality (adjusted hazard ratio (adjHR) for daytime standard deviation 1.16 (95% confidence interval 1.08–1.26)) with no significant differences among age groups. Conversely, systolic BPV was independently associated with cardiovascular mortality only in younger (<50 years) subjects (adjHR for daytime standard deviation 1.72 (95% confidence interval 1.33–2.23)), superseding the predictive value of diastolic BPV in this group. Conclusions: Diastolic short-term BPV independently predicts cardiovascular mortality in hypertensive subjects at all ages, while systolic BPV seems a particularly strong predictor in young adults. If confirmed, these findings might improve the understanding of the prognostic value of BPV, with new perspectives for its possible clinical application.