The geometries and O-17 hyperfine coupling constants in several peroxyl radicals have been determined through tile use of density functional theory. Becke’s three-parameter hybrid exchange functional (B3) together with the correlation functional of Lee, Yang, and Parr (LYP) in combination with a variety of basis sets was used to study basis set effects : Subsequently, the effects of different gradient-correlated functionals were also examined. Results comparable to experimental values are obtained for all of the alkyl peroxyl radicals at the B3LYP level with IGLO-III or s-shell decontracted IGLO-III, 6-311G(d,p), 6-311 + G (2df,p), and the augmented correlation-consistent polarized-valence triple-zeta basis set of D. E. Woon and T, H. Dunning [J, Chem. Phys. 98, 1358 (1993)], R. E. Kendall, T. H. Dunning, and R. J. Harrison [J. Chem. Phys. 96, 6796 (1992)], and T. H. Dunning [J. Chem. Phys. 90, 1007 (1989)]. Calculations imply that the spin density ratio between the inner and outer oxygens is 0.3:0.7, supporting earlier theoretical work [S. L. Boyd, R. J. Boyd, and L. R. C. Barclay, J. Am. Chem. Sec. 112, 5724 (1990)]. Erratic and strongly fluctuating results are exhibited for the fluoroperoxyl radical. Geometries close to the experimental values can be obtained at the B3LYP level, but at the expense of considerable spin contamination. A high degree of spin contamination can also be observed in calculations of the hyperfine coupling constants for this molecule. Possible explanations for the apparent failure to obtain converged results for FOG, apart from the considerable spin contamination, include vibrational, multireference, and matrix effects.