We assess the status of big-bang nucleosynthesis (BBN) in light of the final Planck data release and other recent developments, and in anticipation of future measurements. Planck data from the recombination era fix the cosmic baryon density to 0.9% precision, and now damping tail measurements determine the helium abundance and effective number of neutrinos with precision approaching that of astronomical and BBN determinations respectively. All three parameters are related by BBN. In addition, new high-redshift measurements give D/H to better precision than theoretical predictions, and new Li/H data reconfirm the lithium problem. We present new 7 Be(n, p) 7 Li rates using new neutron capture measurements; we have also examined the effect of proposed changes in the d(p, γ) 3 He rates. Using these results we perform a series of likelihood analyses. We assess BBN/CMB consistency, with attention to how our results depend on the choice of Planck data, as well as how the results depend on the choice of non-BBN, non-Planck data sets. Most importantly the lithium problem remains, and indeed is more acute given the very tight D/H observational constraints; new neutron capture data reveals systematics that somewhat increases uncertainty and thus slightly reduces but does not essentially change the problem. We confirm that d(p, γ) 3 He theoretical rates brings D/H out of agreement and slightly increases 7 Li; new experimental data are needed at BBN energies. Setting the lithium problem aside, we find the effective number of neutrino species at BBN is N ν = 2.86±0.15. Future CMB Stage-4 measurements promise substantial improvements in BBN parameters: helium abundance determinations will be competitive with the best astronomical determinations, and N eff will approach sensitivities capable of detecting the effects of Standard Model neutrino heating of the primordial plasma. 7 Li(p, α) 4 He 7 Li(p, γ) 4 He 4 He Also following [29], we model the uncertainty distribution as a lognormal distribution.This is motivated physically by the idea that the experimental nuclear rates are controlled by several multiplicative factors whose uncertainties thus take this form [17]. In practice the errors are usually sufficiently small that the choice of a lognormal versus Gaussian distribution does not have a large impact on our result.