We examine in detail how BBN theory is constrained, and what predictions it can make, when using only the most model-independent observational constraints. We avoid the uncertainties and model-dependencies that necessarily arise when solar neighborhood D and 3He abundances are used to infer primordial D and 3He via chemical and stellar evolution models. Instead, we use 4He and 7Li, thoroughly examining the effects of possible systematic errors in each. Via a likelihood analysis, we find near perfect agreement between BBN theory and the most model-independent data. Given this agreement, we then assume the correctness of BBN to set limits on the single parameter of standard BBN, the baryon-to-photon ratio, and to predict the primordial D and 3He abundances. For the baryon-to-photon ratio, η, we find a best value, η = 1.8 × 10 −10, corresponding to Ω Bh 2 = 0.0066. We also repeat our analysis including recent measurements of D/H from quasar absorption systems and find that the near perfect agreement between theory and observation of the three isotopes, D, 4He and 7Li is maintained. These results have strong implications for the chemical and stellar evolution of the light elements, in particular for 3He. In addition, our results (especially if the D/H measurements are confirmed) have implications for the stellar depletion of 7Li. Finally, we set limits on the number N v of neutrino flavors, using an analysis which carefully and systematically includes all available experimental constraints. The value N v = 3.0 fits best with BBN and a 95% CL upper limit of N v 4 is established.