Abstract Background: Currently, there is no consensus on the effect of 24-h and 36-h acute total sleep deprivation (ATSD) on human attention. This activation likelihood estimation (ALE) meta-analysis aimed to compare the different patterns of functional magnetic resonance imaging (fMRI activation) between 24-h and 36-h ATSD across attention tasks. Methods: We used GingerALE 2.3.6 software to conduct coordinate-based ALE meta-analysis. The literature related to sleep deprivation, attention, and neuroimaging was searched in four databases: CNKI, PubMed, Web of Science, and PsycINFO from November 1980 to March 2023. Results: We included 16 fMRI-related articles, with 383 participants and 95 foci. The findings revealed that 24-h ATSD and 36-h ATSD may impair different brain areas. After 24-h ATSD, there was significantly reduced brain activation in the parietal-occipital attention lobes and the salience network, including the bilateral superior parietal lobule, right inferior occipital gyrus, and left insula. Increased activation was observed in the sub-lobar regions, including the bilateral thalamus. After 36-h ATSD, there was significantly reduced activation in the frontoparietal attention network, including the left middle frontal gyrus and the right inferior frontal gyrus. Conclusions: This ALE meta-analysis revealed that prolonged ATSD leads to more severe temporary brain damage and a cumulative decrease in the external stimuli captured by humans. This primarily affects the frontal-parietal-occipital attention network and the salience network. Thalamic activation may compensate for dysfunction in the parietal-occipital attention network after 24-h ATSD. Sleep deprivation duration plays a crucial role in the extent of attention impairment.