Research in individuals at clinical high-risk (CHR) for psychosis has focused on subjects with no more than 12 months of present or worsened attenuated positive symptoms. However, the impact of long duration attenuated positive and/or negative prodromal symptoms on outcomes is unclear. Seventy-six CHR subjects with attenuated positive symptoms and at least moderate severity level negative symptoms rated on the Scale of Prodromal Symptoms (SOPS) were prospectively followed for a mean of 3.0 ± 1.6 years. Social and Role functioning was assessed with the Global Functioning: Social and Role scales. Correlations between attenuated positive and negative symptom duration and severity and conversion to psychosis and functional outcomes were analyzed. The average onset of SOPS rated negative symptoms (M = 53.24 months, SD = 48.90, median = 37.27) was approximately twelve months prior to the emergence of attenuated positive symptom (M = 40.15 months, SD = 40.33, median = 24.77, P < 0.05). More severe positive symptoms (P = 0.004), but not longer duration of positive (P = 0.412) or negative (P = 0.754) symptoms, predicted conversion to psychosis. Neither positive symptom duration (P = 0.181) nor severity (P = 0.469) predicted role or social functioning at study endpoint. Conversely, longer negative symptom duration predicted poor social functioning (P = 0.004). Overall, our findings suggest that the severity of attenuated positive symptoms at baseline may be more important than symptom duration for determining individuals at increased risk of developing psychosis. In contrast, long-standing negative symptoms may be associated with persistent social difficulties and therefore have an important position in the treatment of disability.