BACKGROUND.: Cerebral glutamate and gamma-aminobutyric acid (GABA) levels might predict clinical outcome in individuals at ultrahigh risk (UHR) for psychosis but have previously primarily been investigated in smaller cohorts. We aimed to study whether baseline levels of glutamate and GABA in anterior cingulate cortex (ACC) and glutamate in thalamus could predict remission status and whether baseline metabolites differed in the remission versus the nonremission group. We also investigated the relationship between baseline metabolite levels and severity of clinical symptoms, functional outcome, and cognitive deficits at follow-up. METHODS.: About 124 UHR individuals were recruited at baseline. In this, 74 UHR individuals were clinically and cognitively assessed after 12 months, while remission status was available for 81 (25 remission/56 nonremission). Glutamate and GABA levels were assessed at baseline using 3 T proton magnetic resonance spectroscopy. Psychopathology, symptom severity, and remission were assessed with the Comprehensive Assessment of At-Risk Mental States and Clinical Global Impression and functional outcome with the Social and Occupational Functioning Assessment Scale. Cognitive function was estimated with the Cambridge Neuropsychological Test Automated Battery. RESULTS.: There were no differences between baseline glutamate and GABA levels in subjects in the nonremission group compared with the remission group, and baseline metabolites could not predict remission status. However, higher baseline levels of GABA in ACC were associated with clinical global improvement (r = -0.34, N = 51, p = 0.01) in an explorative analysis. CONCLUSIONS.: The variety in findings across studies suggests a probable multifactorial influence on clinical outcome in UHR individuals. Future studies should combine multimodal approaches to attempt prediction of long-term outcome.