Abstract Purpose Intravoxel incoherent motion (IVIM) has been proposed as a means of non-invasive MRI measurement of perfusion parameters such as blood flow and blood volume. Its main competitor in the brain is arterial spin labelling (ASL). In theory, IVIM should not suffer from some of the same limitations as ASL such as poor signal in white matter, and assumptions about arterial arrival times that may be violated in the presence of pathology. Methods In this study we aimed to test IVIM as a viable alternative to ASL for quantitative imaging of perfusion parameters in the brain. First, a direct comparison was performed between IVIM and multi-post label delay pseudo-continuous ASL; second, IVIM images were acquired with and without nulling cerebrospinal fluid; and finally, ultra-high resolution IVIM was performed to minimise partial voluming. Results In all three tests, IVIM failed to disprove the null hypothesis, strongly suggesting that, at least within the brain, the technique does not measure perfusion parameters as proposed. Conclusion Furthermore, the results obtained suggest that the contrast visible in IVIM-derived images is primarily sensitive to cerebrospinal fluid, and not the microvascular blood compartment.