The baryon density of the universe is measured twice, by physics separated by 400,000 years, and the two answers have drifted apart. Big bang nucleosynthesis fixes Omega_b through the primordial deuterium abundance: D/H measured in pristine quasar absorption systems (2.527 +/- 0.030 x 10^-5) combined with nuclear reaction rates yields the baryon density at three minutes. The CMB fixes it independently through the acoustic peak heights at recombination: Planck gives Omega_b h^2 = 0.02237 +/- 0.00015. When the LUNA experiment delivered its precision measurement of the deuterium-burning rate d(p,gamma)3He in 2020, the BBN-inferred value shifted to Omega_b h^2 near 0.02195 +/- 0.00022, opening a gap of about 1.5 to 2 sigma against the CMB, in a quantity that had been advertised as cosmology's cleanest concordance.
The tension is mild in sigma but structurally pointed: both inferences are precision metrology with shrinking error bars, the nuclear input is now laboratory-measured rather than theoretically extrapolated, and the discrepancy moves in a definite direction, with BBN preferring fewer baryons than the CMB. Within ΛCDM there is no freedom: the same Omega_b must serve both epochs, so the gap must resolve into remaining nuclear-rate systematics (the competing theoretical rate calculations differ from LUNA at the relevant energies), deuterium measurement subtleties, or, if it hardens, new physics between BBN and recombination.
The standing is a concordance under strain at the precision frontier: not yet a crisis, but the flagship cross-check of early-universe cosmology no longer agrees as comfortably as the textbooks state, and forthcoming 30-meter-class D/H measurements plus CMB-S4 will push both error bars below the present gap.