In this talk, Dr. Kendrick Kay will describe two recent projects that exploit functional magnetic resonance imaging (fMRI) at ultra-high magnetic field strength (7 Tesla). The first project consisted of whole-brain fMRI retinotopic mapping in 181 healthy adults, as part of the Human Connectome Project (T2*-weighted gradient-echo EPI, 1.6-mm isotropic resolution, 1-s TR, 85 slices, multiband acceleration 5, in-plane acceleration 2). Both group-average and individual-subject results indicate robust signals across much of the brain, including occipital, temporal, parietal, and frontal cortex as well as subcortical areas. The data and resources from this project have been made freely available, and can be used to answer a variety of scientific questions. The second project involved measuring fMRI signals at sub-millimeter resolution during a simple visual experiment (T2*-weighted gradient-echo EPI, 0.8-mm isotropic resolution, 2.2-s TR, 84 slices, multiband acceleration 2, in-plane acceleration 3). Using careful evaluation metrics, Dr. Kay demonstrates that fMRI responses can indeed be measured at sub-millimeter scale with high accuracy and reliability. However, simple inspection of T2*-weighted intensities reveals that these responses, though measurable, are corrupted by a complex network of cortical veins. He shows that venous effects are widespread, heterogeneously distributed, tend to be found in outer cortical depths, and are more prevalent in gyri than sulci. These two projects are strikingly different in measurement protocols and experimental goals. Together, they demonstrate that ultra-high-field fMRI is valuable but requires concerted effort to understand the nature of the measurements and whether valid inferences regarding local neural activity can be made.