To determine if susceptibility to aztreonam could be predicted from cefotaxime or ceftazidime disc diffusion testing, 919 Enterobacteriaceae and 187 Pseudomonas aeruginosa clinical strains were studied. The correlation coefficient between the diameters of inhibition zones was 0.9 for cefotaxime versus aztreonam and ceftazidime versus aztreonam comparisons in Enterobacteriaceae and 0.75 for ceftazidime versus aztreonam comparison in P. aeruginosa. For 99% of the Enterobacteriaceae, there was no risk in predicting susceptibility to aztreonam on the basis of cefotaxime or ceftazidime susceptibility tests. To minimize the risk of the remaining 1% of the strains being erroneously classified as susceptible to aztreonam, ceftoaxime should be tested in preference to ceftazidime, and the production of extended-spectrum beta-lactamases should be tested for using the cefotaxime-clavulanate disc synergy test. For P. aeruginosa strains, susceptibility to aztreonam could be accurately predicted from ceftazidime susceptibility tests for ticarcillin susceptible strains, but for ticarcillin resistant strains, susceptibility to aztreonam should be tested.