The study investigates the amplitude and test-retest reliability of spontaneous low-frequency oscillations (LFOs) in the resting human brain using fMRI. It confirms that gray matter exhibits higher LFO amplitudes than white matter, with the largest amplitudes found in midbrain structures associated with the default-mode network. LFO amplitudes in specific brain regions are reliable across time, and parcellation-based analyses reveal consistent ranking orders of LFO amplitudes among anatomical regions. Detailed examination of individual low-frequency bands shows distinct spatial profiles, with the slow-4 band (0.027–0.073 Hz) showing the most robust amplitudes in the basal ganglia, consistent with findings in awake rats. These results suggest that LFO amplitude measures can contribute to characterizing resting-state fMRI datasets. The study also examines the test-retest reliability of LFO amplitude measures, ALFF and fALFF. ALFF is defined as the total power within the 0.01–0.1 Hz range, while fALFF is the ratio of this power to the total detectable frequency range. Both measures show high reliability in gray matter, with ALFF being more reliable than fALFF. The study finds that LFO amplitudes are more reliable in gray matter than white matter, and that regional differences in LFO amplitudes are consistent across sessions and scanners. The study further explores the distribution of LFO amplitudes across different frequency bands, finding that slow-4 and slow-5 oscillations are primarily detected in gray matter, while slow-3 and slow-2 oscillations are mainly in white matter. These findings suggest that slow-4 and slow-5 oscillations are more relevant to resting-state functional connectivity. The study also finds that slow-4 has higher test-retest reliability and more widespread spatial distribution of reliable voxels than slow-5. The study concludes that LFO amplitudes in gray matter are more reliable and reflect meaningful neuronal activity, while LFO amplitudes in white matter are less reliable and may be influenced by vascular factors. The findings suggest that LFO amplitude measures can be a reliable and robust marker of inter-individual and group differences, with potential applications in clinical and developmental studies. The study also highlights the importance of considering vascular and physiological factors when interpreting LFO amplitude measures.The study investigates the amplitude and test-retest reliability of spontaneous low-frequency oscillations (LFOs) in the resting human brain using fMRI. It confirms that gray matter exhibits higher LFO amplitudes than white matter, with the largest amplitudes found in midbrain structures associated with the default-mode network. LFO amplitudes in specific brain regions are reliable across time, and parcellation-based analyses reveal consistent ranking orders of LFO amplitudes among anatomical regions. Detailed examination of individual low-frequency bands shows distinct spatial profiles, with the slow-4 band (0.027–0.073 Hz) showing the most robust amplitudes in the basal ganglia, consistent with findings in awake rats. These results suggest that LFO amplitude measures can contribute to characterizing resting-state fMRI datasets. The study also examines the test-retest reliability of LFO amplitude measures, ALFF and fALFF. ALFF is defined as the total power within the 0.01–0.1 Hz range, while fALFF is the ratio of this power to the total detectable frequency range. Both measures show high reliability in gray matter, with ALFF being more reliable than fALFF. The study finds that LFO amplitudes are more reliable in gray matter than white matter, and that regional differences in LFO amplitudes are consistent across sessions and scanners. The study further explores the distribution of LFO amplitudes across different frequency bands, finding that slow-4 and slow-5 oscillations are primarily detected in gray matter, while slow-3 and slow-2 oscillations are mainly in white matter. These findings suggest that slow-4 and slow-5 oscillations are more relevant to resting-state functional connectivity. The study also finds that slow-4 has higher test-retest reliability and more widespread spatial distribution of reliable voxels than slow-5. The study concludes that LFO amplitudes in gray matter are more reliable and reflect meaningful neuronal activity, while LFO amplitudes in white matter are less reliable and may be influenced by vascular factors. The findings suggest that LFO amplitude measures can be a reliable and robust marker of inter-individual and group differences, with potential applications in clinical and developmental studies. The study also highlights the importance of considering vascular and physiological factors when interpreting LFO amplitude measures.