The cosmic dipole test is the cosmological principle's most direct examination. Our motion through the universe, 369 km/s as read from the CMB dipole, should imprint a matching kinematic dipole on the counts of distant sources through aberration and Doppler boosting, with an amplitude calculable from the source spectra and number counts (Ellis and Baldwin 1984). If the universe is isotropic on average, the matter dipole and the CMB dipole must agree in both amplitude and direction once local structure is removed.
They do not agree. Wide-area radio and infrared catalogs keep returning source-count dipoles substantially larger than the kinematic expectation: the CatWISE quasar sample shows a dipole roughly twice the predicted amplitude, rejecting the purely kinematic interpretation at 4.9 sigma (Secrest et al. 2021), with earlier radio analyses pointing the same way (Jain and Ralston 1999; Singal 2011). The directions sit near the CMB dipole but not exactly on it, and the excess survives masking, flux cuts, and catalog swaps. Within ΛCDM there is no room for an intrinsic matter dipole at this amplitude: the same observer velocity must produce both dipoles, so a factor-of-two excess requires either correlated systematics across independent surveys at different wavelengths or a genuine large-scale anisotropy the cosmological principle forbids.
The standing is one of the sharpest current challenges to FRW isotropy, with the kinematic interpretation of the CMB dipole itself now under examination. LSST and SKA source counts will push the statistics far past the current samples and resolve the excess's redshift structure.