Current understanding of ultra high risk syndromes for psychosis (UHR) has been based almost entirely on studies of clinical help-seeking populations. The current study aimed to estimate what proportion of the community would meet UHR criteria, to assess whether this was associated with a need for care, and to explore how these individuals relate to those in clinical settings. An epidemiological sample of 208 young adults (aged 18 to 35) was interviewed using the CAARMS (for positive and negative symptoms) and the SPI-A (for basic symptoms), along with measures of functioning and general psychopathology. Help-seeking was measured in relation to both clinical and informal sources of help. Comparisons were also made with a clinical UHR sample from the OASIS service in South London. Thirty individuals met symptomatic criteria for an UHR state (14 met CAARMS criteria, 12 met SPI-A criteria, 4 met both), giving an estimated community prevalence of around 13%. Of these, 66% (n = 20) reported an unmet need for care, 52% (n = 15) had sought some form of help and 33% (n = 9) had engaged in clinical help-seeking. Help-seeking and distress were most associated with negative symptoms and least associated with basic symptoms. Nevertheless, these community UHR subjects were less functionally impaired [t(63) = 3.30, p = .003] and had less severe positive [z = -4.21, p < .001], negative [z = -2.63, p = .017] and general psychopathology [z = -2.74, p = .019] than those already attending clinical services. Results suggest that the UHR criteria can identify something clinically meaningful even in the general population, and that there may currently be individuals who would benefit from outreach by existing UHR services. However, results also suggest that the current focus on positive symptoms may be insufficient for identifying those in need of care.