Ongoing neural activity fluctuates over a broad range of timescales. Variations in intrinsic timescales across the forebrain relate to the functional specialization of cortical areas along the visual hierarchy [1, 2]. Furthermore, these timescales adapt to task demands, for example, during selective attention [3] and working memory [4] tasks. However, the distribution of timescales and their relationship to brain structure and behavior have not been characterized beyond the forebrain, despite the evolutionary relevance of midbrain and hindbrain areas. These areas are preserved across different species suggesting their potential relevance for survival. Here, we measured timescales in the spontaneous spiking activity of single neurons using a brain-wide dataset of 10-minute Neuropixels recordings in mice [5]. Data consists of 27,960 neurons from 308 brain regions covering the forebrain, midbrain, and hindbrain areas. We showed that autocorrelation of individual neurons’ activity can be described by up to 4 different timescales, ranging from tens of milliseconds to several seconds. To compare timescales across different neurons, we defined an effective timescale as the weighted sum of all timescales for each neuron. Compatible with previous findings [2], we observed a hierarchical organization of timescales in the cortex, but not in the thalamus (Fig 1a). Surprisingly, we found infra-slow intrinsic timescales in the midbrain and hindbrain areas, much slower than the timescales in the forebrain (Fig 1b). Moreover, we tested the relationship between brain-wide heterogenous timescales and molecular signatures of cell types in gene expression profiles. We used spatial expression profiles of 4,345 genes from the Allen Gene Expression Atlas to predict the local intrinsic timescales across the entire brain. The gene-expression profiles accounted for 6% of the variance in the spatial distribution of timescales, significantly more than the variance explained by the brain-region parcellation. Finally, we studied the relevance of intrinsic timescales to the perceptual decision-making task that the mice performed. Our results suggest that neurons that code for choice and feedback have generally slower timescales (Fig 1c) but this relationship is not present for coding the visual stimulus. Overall, we provide a comprehensive brain-wide map of intrinsic timescales for the mouse brain and demonstrate their relation to gene expression and behavior.