The sensitivity to subject motion is one of the major challenges in functional MRI (fMRI) studies in which a precise alignment of images from different time points is required to allow reliable quantification of brain activation throughout the scan. Especially the long measurement times and laborious fMRI tasks add to the amount of subject motion found in typical fMRI measurements, even when head restraints are used. In case of moving subjects, prospective motion correction can maintain the relationship between spatial image information and subject anatomy by constantly adapting the image slice positioning to follow the subject in real time. Image-based prospective motion correction is well-established in fMRI studies and typically computes the motion estimates based on a volume-to-volume image registration, resulting in low temporal resolution. This study combines fMRI using simultaneous multislice imaging with multislice-to-volume-based image registration to allow sub-TR motion detection with subsequent real-time adaption of the imaging system. Simultaneous multislice imaging is widely used in fMRI studies and, together with multislice-to-volume-based image registration algorithms, enables computing suitable motion states after only a single readout by registering the simultaneously excited slices to a reference volume acquired at the start of the measurement. The technique is evaluated in three human BOLD fMRI studies (n = 1, 5, and 1) to explore different aspects of the method. It is compared to conventional, volume-to-volume-based prospective motion correction as well as retrospective motion correction methods. Results show a strong reduction in retrospectively computed residual motion parameters of up to 50% when comparing the two prospective motion correction techniques. An analysis of temporal signal-to-noise ratio as well as brain activation results shows high consistency between the results before and after additional retrospective motion correction when using the proposed technique, indicating successful prospective motion correction. The comparison of absolute tSNR values does not show an improvement compared to using retrospective motion correction alone. However, the improved temporal resolution may provide improved tSNR in the presence of more exaggerated intra-volume motion.