Objective: The physiologic response to exercise may provide valuable prognostic information. We investigated the association of blood pressure (BP) measurements during exercise stress testing (EST) with long-term risk of myocardial infarction, stroke or death (major adverse cardiovascular event, MACE), as well as the development of new-onset hypertension. Methods: A retrospective analysis of treadmill ESTs (years 2005–2019) performed by the Bruce protocol in patients aged 35–75 years without a history of cardiovascular disease ( n = 14 792; 48% women). BP was documented at rest, submaximal exercise (Bruce stage-2), peak exercise and recovery (2 min). Association of SBP measures with study outcomes during median follow-up of 6.5 years was investigated. Results: Highest vs. lowest SBP quartile at rest (≥140 vs. <120 mmHg), submaximal-exercise (≥170 vs. <130 mmHg), peak-exercise (≥180 vs. ≤145 mmHg) and recovery (≥160 vs. <130 mmHg) was associated with an increase in the adjusted hazard ratio and 95% confidence interval (CI) for MACE: 1.53 (1.23–1.88), 1.33 (1.01–1.76), 1.30 (1.05–1.61), 1.35 (1.09–1.68), respectively. The association between SBP at submaximal exercise and recovery with MACE displayed a J-shaped pattern. Among nonhypertensive patients ( n = 8529), excessive SBP response to peak exercise (≥190 mmHg in women and ≥210 mmHg in men) was an independent predictor of hypertension [hazard ratio (95% CI)]: 1.87 (1.41–2.48), as were SBPs during submaximal exercise [>160 vs. ≤130 mmHg: 2.44 (1.97–3.03)] and recovery [≥140 vs. ≤120 mmHg: 1.65 (1.37–1.98)]. Conclusion: BP measurement during rest, exercise and recovery phases of EST provides incremental prognostic information regarding long-term risk for cardiovascular events and the probability for developing hypertension.