The assumption of a dynamic coupling between regional cerebral blood flow (rCBF) and cerebral glucose metabolic rates (rCMRGlu) has been challenged by simultaneous measurements of both. Through the use of a dual-headed gamma camera with a 511-keV collimator applying the double isotope 18F-FDG and 99mTc-HMPAO SPECT technique, the uptake rates of these isotopes can be semi-quantitatively evaluated. Sixteen depressed patients, diagnosed by ICD-10 criteria and assessed with the 17-item Hamilton Rating Scale for Depression (HRSD), were studied. Based on the severity of HRSD-rated anxiety (item 10: low=1-21; high=3-4), two eight-patient subgroups were formed and compared with 12 age- and handedness-matched healthy control subjects. As regions of interest, we selected areas implicated in the neuroanatomy of anxiety and depression: hippocampus (hippo), basal ganglia (BG) and gyri temporales superiores (G.t.s.). In the control subjects, a significant statistical coupling between rCBF and rCMRGlu was revealed by the Spearman correlation coefficient only in left hippo and left BG. Patients in the low-anxiety subgroup demonstrated a marked dynamic coupling bilaterally for the G.t.s., while patients in the high-anxiety subgroup showed a significant statistical correlation of rCBF and rCMRGlu only in the left G.t.s. These findings indicate that a dynamic coupling between blood flow and glucose metabolism exists only in distinct brain regions, and that the depressive illness has an uncoupling effect on this correlation in the left BG. Furthermore, our results suggest that the HRSD anxiety score might interact with the underlying depressive illness to influence the relationship of rCBF and rCMRGlu.