Gradient-corrected density functional theory calculations using the B3LYP hybrid functional have been performed on Co(acacen)(pyridine) and Co(salen)(pyridine) complexes and their adducts with molecular oxygen (acacen = 2,11-dihydroxy-4,9-dimethyl-5,8-diaza-2,4,8,10-dodecatetraene, salen = 1,6-bis(2-hydroxyphenyl)-2,5-diaza-1,5-hexadiene). Calculated optimized geometries and dioxygen binding energies show good agreement with experiment. The binding energy for Co(acacen)(pyridine) is calculated to be 53.5 kJ mol(-1) (experiment = 63.1 kJ mol(-1)), and the binding energy for Co(salen)(pyridine) is calculated to be 46.8 kJ mol(-1) (experiment = 51.8 kJ mol(-1)). The calculated O-O distances are in the range 1.28-1.29 Angstrom, which compares to 1.26 Angstrom from crystal structure data for Co(acacen)(pyridine)(O-2). Model complexes for Co(acacen)(pyridine) are also examined, and ligand substitution effects correlated with the computed Co-O distance. The calculated O-O vibrational frequency (1243 cm(-1)) is slightly higher than typical experimental frequencies while the calculated Co-O frequency is 439 cm(-1). The effect of basis set variation on the computed binding energy in the model complexes was also examined, and it was found that a more flexible basis set was required when using standard effective core potentials for cobalt. In addition, the inclusion of basis set superposition error and the zero point correction were found to severely reduce the magnitude of the calculated binding energy.