Caffeine intake before adenosine stress myocardial perfusion imaging may cause false negative findings. We hypothesized that the antagonistic effect of caffeine can be measured by T1 relaxation times in rest and adenosine stress cardiac magnetic resonance imaging (CMR), as T1 mapping techniques are sensitive to changes in myocardial blood volume. We prospectively analyzed 105 consecutive patients with adenosine stress perfusion CMR on a 1.5-T MRI system. Rest and stress T1 mapping was performed using Modified Look-Locker Inversion recovery. T1 reactivity was defined as difference in T1(rest) and T1(stress) (a dagger T1). Fifteen patients drank coffee within 4 h of CMR (<4H caffeine group), and 10 patients had coffee the day before (> 8H caffeine group). Comparison was made to patients without self-reported coffee intake: 50 with normal CMR (control group), 18 with myocardial ischemia, and 12 with myocardial infarction. The national review board approved the study; all patients gave written informed consent. The <4H caffeine group showed inverted a dagger T1 of -7.8 % (T1(rest) 975 +/- 42 ms, T1(stress) 898 +/- 51 ms, p <0.0005). The > 8H caffeine group showed reduced T1 reactivity (1.8 %; T1(rest) 979 ms, T1(stress) 997 ms) compared to the controls (4.3 %; T1(rest) 977 +/- 40 ms, T1(stress) 1018 +/- 40 ms), p <0.0005. Ischemic and infarcted myocardium showed minimal T1 reactivity (0.2 and 0.3 %, respectively). Caffeine intake inverts the adenosine effect during stress perfusion CMR as measured by T1 mapping. T1 reactivity can assess the adequacy of adenosine-induced stress in perfusion CMR.