Psychosis is known to exist on a continuum in the population, ranging from infrequent, subclinical psychotic-like experiences to full-blown psychotic disorders. Resting-state functional connectivity studies report widespread dysconnectivity at various points along the psychosis continuum, yet a comprehensive mapping of psychosis in the human connectome remains elusive. This dissertation aims to build a resting-state functional connectivity model reproducible and generalizable across the psychosis continuum with a systematic approach to large community and clinical samples. Measurement properties of the human connectome derived with various methods were first compared to identify the human connectome with optimal test-retest reliability. The identified human connectome was then used to build a cross-validated model for psychotic-like experiences in a large community sample (N = 855). Lastly, this model was validated in a clinical sample with patients with psychosis and first-degree relatives. Findings suggest that independent component analysis with dimensionalities above 100 yielded human connectomics with optimal reliability. A model involving primarily connections in the frontoparietal, default, cingulo-opercular, and dorsal attention networks explained as much as 3.4% of variance in psychotic-like experiences in healthy adults. In the clinical sample, model score explained psychoticism and schizotypy across patients with psychosis, first-degree relatives, and healthy controls (partial correlation ranged from 0.19 to 0.51). Findings provide direct evidence for the psychosis continuum encoded in the human connectome. The quantifiable resting state functional connectivity model facilitates validation in additional samples and shows preliminary potential for clinical utility.