Culturable counts of antibiotic resistant, genetically engineeredPseudomonas fluorescens were determined on antibiotic-containing plate count agar during starvation in water. Prior to starvation, colony counts obtained on all media separated into two groups. The mean of the colony counts on plate count agar with or without tetracycline (4.9 × 10⁶ ml⁻¹) was significantly higher than the mean colony counts on plate count agar containing either nalidixic acid or nalidixic acid plus tetraclycline (2.5×10⁶ ml⁻¹). After 20 days of starvation the highest mean colony counts continued to be obtained on plate count agar (7.2 × 10⁶ ml⁻¹) with slightly, but significantly, lower counts obtained on plate count agar containing either nalidixic acid (5.6 × 10⁶ ml⁻¹) or tetraclycline (1.5×10⁶ ml⁻¹). A combination of nalidixic acid and tetracycline in plate count agar, however, dramatically reduced colony counts (8.3 × 10² ml⁻¹) after this starvation period. The addition of catalase to plate count agar containing nalidixic acid and tetracycline negated the effect caused by this combination of antibiotics. When colony counts obtained over the entire 20 day incubation were considered, the addition of MgSO₄ to plate count agar containing nalidixic acid and tetracycline resulted in a significant increase in colony counts. Other combinations of antibiotics, nalidixic acid+carbenicillin, nalidixic acid+kanamycin, streptomycin+tetracycline, streptomycin+carbenicillin, rifampicin+tetracycline, rifampicin+carbenicillin, and rifampicin+kanamycin, did not inhibit colony formation of starved cells. Antibiotic resistant strains ofP. putida andEscherichia coli also displayed sensitivity to the combination of nalidixic acid and tetracycline in plate count agar after starvation.