Accurate thermochemical data for compounds containing C/H/N/O are required to underpin kinetics simulation and modeling of the reactions of these species in different environments. There is a dearth of experimental data so computational quantum chemistry has stepped in to fill this breach and to verify whether particular experiments are in need of revision. A number of composite model chemistries (CBS-QB3, CBS-APNO, G3, and G4) are used to compute theoretical atomization energies and hence enthalpies of formation at 0 and 298.15 K, and these are benchmarked against the best available compendium of values, the Active Thermochemical Tables or ATcT. In general the agreement is very good for some 28 species with the only discrepancy being for hydrazine. It is shown that, although individually the methods do not perform that well, collectively the mean unsigned error is <1.7 kJ mol(-1); hence, this approach provides a useful tool to screen published values and validate new experimental results. Using multiple model chemistries does have some drawbacks but can produce good results even for challenging molecules like HOON and CN2O2. The results for these smaller validated molecules are then used as anchors for determining the formation enthalpies of larger species such as methylated hydrazines and diazenes, five- and six-membered heterocyclics via carefully chosen isodesmic working reactions with the aim of resolving some discrepancies in the literature and establishing a properly validated database. This expanded database could be useful in testing the performance of computationally less-demanding density function methods with newer functionals that have the capacity to treat much larger systems than those tested here.