A once-daily dosage regimen has been recently recommended in the use of aminoglycoside antibiotics since they induce a postantibiotic effect. In choosing this regimen, one must determine the most appropriate time of day for administration of the drug. We investigated the effects of the timing of amikacin (AMK) administration on the kinetics, the efficacy against intraperitoneal infection with Pseudomonas aeruginosa, and the toxicity of AMK in mice with and without immunosuppression. We found circadian variations in the kinetics, efficacy, and toxicity of the drug in mice. Male and female ICR mice, which were housed under a light-dark (12:12 h) cycle with free food and water intake, were injected subcutaneously with AMK sulfate 50 mg/kg body wt. There was a circadian variation in AMK clearance for both sexes with the maximum value in the dark phase and the minimum in the light phase after a single administration. When AMK 500 mg/kg/day was repeatedly administered once daily for 30 days, higher toxicity was demonstrated in mice injected with the drug at the time of day with lower AMK clearance, although no difference was demonstrated in the toxicity between the two time points with different AMK clearance when AMK 1,500 mg/kg was administered in a single dose. The ED50 of AMK to cure the infected mice in the midlight phase (13:00 h) with lower clearance was significantly lower than that in the middark phase (01:00 h) with higher clearance. In contrast, the ED50 in the early light phase (09:00 h) was significantly lower than that in the early dark phase (21:00 h), although AMK clearance was no different between these two different time points. In mice premedicated with cyclophosphamide to suppress immune functions, the difference in the ED50 of AMK was still demonstrated between 13:00 and 01:00 h, but not between 09:00 and 21:00 h. The present study shows not only that there were circadian variations in both AMK clearance and toxicity after repeated administration, but also that there was a circadian variation in the efficacy of AMK in mice infected with P. aeruginosa. These results suggest that the timing of drug administration should be considered in pharmacotherapy with AMK and that the most appropriate time of administration in mice and nocturnal animals may be in the midlight (resting) phase. They also suggest that the ED50 of AMK against P. aeruginosa infection may be influenced not only by the circadian variation in pharmacokinetics but also by the variations in immune systems suppressed by cyclophosphamide.