Echocardiographic studies have suggested an association between diastolic dysfunction and exercise intolerance. The aim of this study was to examine the relationship between exercise capacity and left ventricular (LV) function during stress myocardial scintigraphy, and to investigate whether or not this relationship is caused by ischemia during exercise.The studied patients underwent technetium-99m sestamibi quantitative gated SPECT, including treadmill exercise. Myocardial stress images were acquired 30 min after the first tracer injection (370 MBq) during maximal exercise. Three hours later, the second tracer (740 MBq) was injected, and resting images were acquired 30 min after this injection. The presence of ischemia was determined by tracer accumulation. From the same data source, LV diastolic parameters [first third filling fraction (1/3FF), first third filling rate (1/3FR), peak filling rate (PFR) and time to PFR (TPF)], and systolic parameters [ejection fraction (EF), peak ejection rate (PER), time to PER (TPE) and first third ejection fraction (1/3EF)] were analyzed.Subjects with exercise inability (<6 METs) were excluded. In 45 patients, diastolic parameters 1/3FF, 1/3FR, PFR and TPF correlated significantly with exercise duration (r = 0.32*, 0.37*, 0.37* and -0.40(#), respectively; *p < 0.05, (#) p < 0.01), but systolic parameters EF, PER, TPE and 1/3EF did not. At rest, 1/3FF, PFR and PER were significantly increased, suggesting functional deterioration during exercise. Even after 3 h, 1/3FR, PFR and TPF still correlated significantly with exercise duration (r = 0.29*, 0.36* and -0.30*, respectively; *p < 0.05). Such findings were observed even when the 10 patients who exhibited ischemia during exercise were excluded (1/3FR: r = 0.34*; PFR: r = 0.37*; TPF: r = -0.36*; *p < 0.05, n = 35).Our findings suggested that LV diastolic dysfunction, not systolic dysfunction, is associated with limited exercise capacity independent of the occurrence of ischemia.