A total of 22 substituted benzene radicals are investigated by means of density functional theory (DFT). Two different functional schemes, the B3LYP hybrid DFT-HF functional and the ‘pure’ gradient corrected DFT functional PWP86, previously employed successfully in calculations of radical hyperfine parameters, are employed. The results are compared with experimental data. The same strategy is used for all calculations: initial geometry optimization using a double-ζ basis set, followed by single point hyperfine calculations using the larger IGLO-III, 6-311G(2df, p) and aug-cc-pCVTZ basis sets. It is found that of the two functionals, the PWP86 functional gives the more homogeneous data, with nearly constant underestimations in HFCCs by 5–15%. With the B3LYP approach the smaller couplings, found at the meta positions, generally are overestimated considerably. The best overall performance is observed at the PWP86/6-311G(2df, p) level.