Objectives: Impaired attention and language functions are common in psychosis, but have been less explored in subjects with ultra-high risk for psychosis (UHR). The aim of the study was to investigate differences in language lateralization and auditory attention in UHR subjects compared to healthy controls with a dichotic listening paradigm. In addition, symptoms from The Structural Interview for Prodromal Syndromes (SIPS) were explored in relation to performance on dichotic listening. Methods: The UHR subjects (n = 46, female = 28, mean age = 17.9) were compared to a group of healthy controls (n = 40, female = 20, mean age = 16.8). A split-plot repeated measures analysis of covariance was conducted with group as between-subjects factor and attention conditions (non-forced, forced-right, forced-left) and side (right ear, left ear) as repeated measures factors (2×3×2 design) using gender, age and handedness as covariates. SIPS symptoms were subjected to Spearman's r correlations with laterality indexes and attentional gain in each ear. Results: There was a statistically significant three-way interaction of group (UHR, healthy controls) × forced condition (non-forced, forced-right, forced-left) × side (right ear, left ear), p = 0.048. The effect was due to an interaction between group × side in the forced-left condition. There were no significant differences between UHR subjects and healthy controls in the non-forced condition. Right ear gain correlated with "Perceptual abnormalities/Hallucinations" (P4), r = 0.486, p = 0.001. Conclusion: UHR subjects demonstrated impairment in top-down attentional mechanisms, but showed no language lateralization abnormalities. Impairment in top-down attentional mechanisms are frequently reported from dichotic listening studies in patients with schizophrenia. Higher levels of perceptual abnormalities and hallucinatory experiences were associated with enhanced report from the right ear in the forced-right condition.