Synergism between two antibiotics is usually tested by a checkerboard titration technique, or by time-kill methods. Both methods have the disadvantage that synergism is determined at constant concentrations of the antibiotics, which do not reflect reality in vivo. In the present study we determined whether synergism between tobramycin and ceftazidime can be found at declining concentrations below the MIC, and whether change in dosing sequence of the antibiotics would result in differences in killing. Three monotherapy and six combination therapy schedules were tested in an in vitro pharmacokinetic model, using a Pseudomonas aeruginosa resistant to both antibiotics. During all q8h dosing schedules the peak concentration (Cmax) was adjusted to the MIC for the strain of both antibiotics. During all monotherapy regimens bacterial growth was present, while all six combination therapy schedules showed significant killing. At t = 24 h there were no differences between all combination therapy schedules, but at t = 8 h the two combination therapy schedules with administration of tobramycin once daily showed a significantly faster killing. By using the area under the killing curve (AUKC) as a parameter for synergistic killing, simultaneous combination therapy starting with tobramycin once daily was significantly better than all other regimens. We conclude that there is synergism between tobramycin and ceftazidime at declining antibiotic concentrations below the MIC, resulting in a pronounced killing of a resistant Pseudomonas strain. Infections due to resistant Pseudomonas strains could possibly be treated by a synergistic combination of these drugs.