Recently, Gerstenberg et al. reported that there is a therapeutic threshold for the steady-state plasma concentrations of fluvoxamine for depressed patients.1 However, the levels necessary to treat bulimia nervosa remain unclear. In the present study we reported six patients with bulimia nervosa who responded to fluvoxamine, and for whom we performed therapeutic drug monitoring regarding plasma fluvoxamine levels. Six women diagnosed with bulimia nervosa (DSM-IV) and who were being treated with fluvoxamine were enrolled in the study. They ranged in age from 14 to 37 years old. None had received any antidepressants, antipsychotics, or mood stabilizers within a week prior to the study. All patients were given fluvoxamine on a flexible-dose regimen. Each patient performed a self-evaluation using the Bulimic Investigatory Test–Edinburgh (BITE);2 and the Hamilton Rating Scale for Depression (HAM-D) was rated by one experienced psychiatrist (AI-S). The plasma fluvoxamine concentrations of each patient were also measured weekly. Clinical improvement using BITE was assessed 8 weeks after the treatment with fluvoxamine. We defined the response as BITE score <20 points. The protocol of the present study was approved by the Ethics Committee of the University of Occupational and Environmental Health. Written informed consent was obtained from all subjects. The initial fluvoxamine dose was 25–50 mg/day, which was increased to 150–200 mg/day. None of the patients responded to treatment with fluvoxamine within 4 weeks after the treatment was begun, but all patients responded by 8 weeks. The average duration to response was 6.2 ± 1.4 weeks. The plasma fluvoxamine concentrations and fluvoxamine dosage at the point (by 8 weeks into treatment) when all six subjects responded to fluvoxamine ranged from 98 ng/mL and 150 mg/day, to 262 ng/mL and 200 mg/day, respectively. The HAM-D scores did not alter before (10 ± 4) or 8 weeks after (7 ± 5) fluvoxamine treatment. We have presented six cases of bulimia nervosa in which the patients did not respond at 4 weeks but responded at 8 weeks and tolerated treatment well with fluvoxamine. We found that the lowest level of plasma fluvoxamine for a response to occur was 98 ng/mL for these bulimic patients. We previously reported a linear relationship between dosages of fluvoxamine and plasma fluvoxamine concentrations.3 However, there is a variation in plasma fluvoxamine levels between individuals when the same dosage of fluvoxamine is administered.4 Interestingly, HAM-D scores in the six patients did not change after fluvoxamine treatment. Therefore, enhancement of the serotonin system by fluvoxamine is related to recovering from the bulimia, which is independent of relieving the depressive symptoms of the patients. Further research to compare patients who started to respond to the dose at 4 weeks with those who started to respond to the dose at 8 weeks is needed to confirm this preliminary result.