In a recent letter (J. Phys. Chem. A, 2001, 105, 1), we argued that, although all major thermochemical tables recommend a value of DeltaHdegrees(f0)(OH) based on a spectroscopic approach, the correct value is 0.5 kcal/mol lower as determined from an ion cycle. In this paper, we expand upon and augment both the experimental and theoretical arguments presented in the letter. In particular, three separate experiments (mass-selected photoionization measurements, pulsed-field-ionization photoelectron spectroscopy measurements, and photoelectron-photoion coincidence measurements) utilizing the positive ion cycle to derive the O-H bond energy are shown to converge to a consensus value of the appearance energy AE(0)(OH+/H2O) = 146117 +/- 24 cm(-1) (18.116(2) +/- 0.003(0) eV). With the most accurate currently available zero kinetic energy photoionization value for the ionization energy IE(OH) = 104989 +/- 2 cm(-1), corroborated by a number of photoelectron measurements, this leads to D-0(H-OH) = 41128 +/- 24 cm(-1) = 117.59 +/- 0.07 kcal/mol. This corresponds to DeltaH(f0)(OH) = 8.85 +/- 0.07 kcal/mol and implies D-0(OH) = 35593 +/- 24 cm(-1) = 101.76 +/- 0.07 kcal/mol. These results are completely supported by the most sophisticated theoretical calculations ever performed on the HxO system, CCSD(T)/aug-cc-pVnZ, n = Q, 5, 6, and 7, extrapolated to the CBS limit and including corrections for core-valence effects, scalar relativistic effects, incomplete correlation recovery, and diagonal Born-Oppenheimer corrections. These calculations have an estimated theoretical error of less than or equal to0.2 kcal/mol based on basis set convergence properties. They reproduce the experimental results for dissociation energies, atomization energies, and ionization energies for the HxO system to within 0.0-0.2 kcal/mol. In contrast, the previously accepted values of the two successive bond dissociation energies of water differ from the current values by 0.5 kcal/mol. These values were derived from the spectroscopic determinations of D-0(OH) using a very short Birge-Sponer extrapolation on OH/OD A(1)Sigma(+). However, on the basis of a calculation of the A state potential energy curve (with a multireference single and double excitation wave function and an aug-cc-pV5Z basis set) and an exhaustive reanalyzis of the original measured data on both the A and B states of OH, the Birge-Sponer extrapolation can be demonstrated to significantly underestimate the bond dissociation energy, although only the last vibrational level was not observed experimentally. The recommended values of this paper affect a large number of other thermochemical quantities which directly or indirectly rely on or refer to D-0(H-OH), D-0(OH), or DeltaHdegrees(f)(OH). This is illustrated by an analysis of several reaction enthalpies, deprotonation enthalpies, and proton affinities.