Magnetic susceptibility-based methods are emerging cardiovascular MR (CMR) techniques for myocardial tissue characterization [1, 2]. The linear relationship between magnetic field strength and microscopic susceptibility effects [3], renders susceptibility weighted MRI at ultrahigh magnetic fields (B₀ ≥ 7.0T) (UHF) conceptually appealing to pursue high spatial resolution temporally resolved T₂^* mapping [1]. Since T₂^* is highly dependent on main magnetic field (B₀) homogeneity [4], meaningful interpretation of these results requires temporally resolved B₀ assessment. Shah et al. [5] reported temporal variation of the main field to be negligible across the cardiac cycle at 1.5T, but B₀ inhomogeneities are increased at UHF [3, 6]. For this reason, this pioneering study investigates B₀ variation in the myocardium over the cardiac cycle at 7.0T and elucidates its implications for myocardial T₂^* mapping in healthy volunteers.Three healthy volunteers were scanned using a 7.0T whole body MR system (Siemens Healthcare, Erlangen, Germany). A 16 channel transceiver array tailored for CMR at 7.0T was used for signal excitation/reception [7]. Volume selective B₀ shimming adjusted to the heart was applied prior to T₂^* weighted acquisitions. CINE T₂^* and B₀ mapping was carried out in mid-ventricular short axis (SAX) views employing a segmented, cardiac gated, breath-held, multi-shot multi-echo gradient echo technique [1] (spatial resolution=(1.4x1.4x4.0)mm³, TE=2.04-10.20ms, ΔTE=1.02ms) [1]. CINE four chamber views were acquired similarly to measure the through-plane B₀ variation (TEs=3.06, 7.14ms, spatial resolution=(2.8x2.8x4.0)mm³). An MR stethoscope (MRI.TOOLS GmbH, Berlin, Germany) was used for cardiac triggering. Field maps were calculated offline in MATLAB (The Mathworks, Natick, MA) using a phase difference method [8]. The B₀ maps were filtered using a Gaussian low-pass to reduce high frequency noise contributions while maintaining macroscopic B₀ variations. Since susceptibility contrast is determined by intravoxel field gradients rather than by overall changes in B₀, it is essential to investigate the change of these gradients. Intravoxel field gradients for mid-ventricular short axis views were calculated based on in-plane field variations together with through-plane gradients estimated from B₀ profiles in the septum of the CINE four chamber views. Septal segments (8 and 9, [9]) are less prone to susceptibility artifacts than other myocardial segments [3] and hence commonly assessed in clinical routine. Therefore mean septal intravoxel gradients were analyzed for all cardiac phases. Finally, the expected T₂^* variations caused by macroscopic field gradients were estimated using an analytic approach. The post-processed data was averaged for all volunteers, while interpolation based on R-R interval duration was used to match cardiac phases of different volunteers.Figure 1 illustrates B₀ field in-plane and through-plane maps of one volunteer for 6 out of 19 phases distributed along the cardiac cycle. Through-plane B₀ gradients (mean=0.4±0.1Hz/mm; mean temporal range (max-min)=0.9±0.2Hz/mm) estimated from septal profiles in four chamber views (Figure 1 bottom) were small compared to in-plane septal gradients (mean=2.3±0.5Hz/mm; mean range=0.6±0.1Hz/mm). Intravoxel gradient maps are shown in Figure 2. Intravoxel gradients were comparable for all cardiac segments with exception of the antero- and inferolateral segments which showed higher values (compare Figure 2 center). End-systolic and end-diastolic intravoxel gradients, T₂^* and estimated gradient-induced ΔT₂^* maps of one volunteer are shown in Figure 3A. Figure 3B illustrates the intravoxel gradient evolution across the cardiac cycle averaged over all volunteers and its effect on T₂^*. The mean range (max-min) of the septal intravoxel B₀ gradients was 1.9±0.5Hz throughout the cardiac cycle, while the mean gradient was ranging from 2.8-4.5Hz (black plot). Mean septal T₂^* was found to be 16.1±1.0ms (blue plot). The mean macroscopic gradient-induced ΔT₂^* (red plot) was estimated as -1.0±0.2ms and was varying in a range (max-min) of 1.0±0.6ms, which is small compared to the mean range of T₂^* changes found 3.8±1.2ms.This study investigated macroscopic B₀ variation in the heart with focus on the intraventricular septum over the cardiac cycle at 7.0T and its effects on T₂^* mapping. B₀ field variation was found to induce a nearly constant but minor offset in septal T₂^*. Consequently, macroscopic magnetic field variation across the cardiac cycle can be considered to be negligible for septal T₂^* mapping. This result provides encouragement for temporally resolved susceptibility sensitized CMR at UHF. Additionally, the T₂^* offset in each phase is small compared to their absolute value indicating that septal T₂^* mapping is reliable even at UHF. In conclusion, macroscopic B₀ inhomogeneities and their temporal changes in the septum were found to be minor and the feasibility of septal dynamic T₂^* mapping at 7.0 T was demonstrated.