The standard cosmological model Lambda CDM is described by the Friedman-Lemaitre-Robertson-Walker (FLRW) metric, which requires that the universe be isotropic and homogeneous on large scales, an assumption called the Cosmological Principle. If this assumption is accurate, then the dipole anisotropy observed in the cosmic microwave background (CMB) corresponds to our motion with respect to large-scale structure at approximately 370 km s(-1), which can be tested by measuring the corresponding dipole predicted in counts of cosmologically distant sources. This consistency test, first proposed in 1984 by Ellis & Baldwin, became possible in the twenty-first century with the advent of large catalogues of radio sources and quasars. Subsequent Ellis-Baldwin tests have consistently shown an anomalously large dipole, two to three times larger than predicted by the kinematic interpretation of the CMB dipole, which has recently reached a statistical significance of over 5 sigma . In these proceedings, I review the Ellis-Baldwin test, the key results that revealed this anomaly, and comment on the status of research on this problem, which threatens a foundational assumption underpinning FLRW-based cosmologies such as Lambda CDM. This article is part of the discussion meeting issue 'Challenging the standard cosmological model'.