Clarifying dynamic fluctuations in resting-state connectivity in individuals at risk for psychosis (termed clinical high risk [CHR]) may inform understanding of psychotic disorders, such as schizophrenia, which have been associated with dysconnectivity and aberrant salience processing. Dynamic functional connectivity (DFC) investigations provide insight into how neural networks exchange information over time. Currently, there are no published DFC studies involving CHR individuals. This is notable, because understanding how networks may come together and disassociate over time could lend insight into the neural communication that underlies psychosis development and symptomatology. A sliding-window analysis was utilized to examine DFC (defined as the standard deviation over a series of sliding windows) in resting-state scans in a total of 31 CHR individuals and 28 controls. Clinical assessments at baseline and 12 months later were conducted. CHR participants exhibited less DFC (lower standard deviation) in connectivity involving areas of both the salience network (SN) and default mode network (DMN) with regions involved in sensory, motor, attention, and internal cognitive functions relative to controls. Within CHR participants, this pattern was associated with greater positive symptoms 12 months later, possibly reflecting a mechanism behind aberrant salience processing. Higher SN−DMN internetwork DFC related to elevated baseline negative symptoms, anxiety, and depression in CHR participants, which may indicate neurological processes underlying worry and rumination. Overall, through highlighting unique DFC properties within CHR individuals and detecting informative links with clinically relevant symptomatology, results support dysconnectivity and aberrant salience processing models of psychosis.