The number of patients with internet addiction disorder (IAD), especially among school-age children, is on the rise. Development of objective examination technique to assist the current diagnostic methods using medical interview and inquiry tests is desirable for detection of IAD at its early stage. In this study, we extracted the values of functional connectivity (FC) that correlated with a tendency of IAD, using resting-state functional magnetic resonance imaging (rs-fMRI) data. We recruited 40 males [mean age (SD): 21.9 (0.9) years] with no neurological disorders, conducted rs-fMRI recordings for 7 min 30 s, and administered five questionnaires including internet addiction test (IAT), to assess their mental states. The IAT scores of all participants were in the range of healthy to a tendency of IAD. The FC values were calculated using the cross-correlation of time-series signals in a low frequency band (0.017 to 0.09Hz) between all possible connections of brain region pairs defined by Automated Anatomical Labeling (AAL). For each region pair, we calculated Pearson's correlation coefficients “γ” between the FC values and the scores of IAT as well as other inventories for psychological states, in all subjects. The “γ” were subjects-cross-validated, and the FC of the survived region pairs were statistically assessed by correction of multiple comparison. Finally, we obtained some region pairs in which FC specifically correlated with the IAT scores, but not with other psychological scores. These pairs had mostly negative “γ” over long-distance connections across right and left brain hemispheres. The results suggested that the functional connectivity between specific brain regions was significantly degraded already at the stage prior to the onset of IAD. We expect that our connectivity method can be an objective tool for detecting a tendency of IAD to assist the present diagnostic methods.