Researching from a symptom approach avoids possible spurious associations, given the co-occurrence of symptoms in a disorder (Costello, 1992). Here, we deepen the evidenced relationships among anxiety, delusions and hallucinations.

We intended to assess differences in Anxiety Sensitivity dimensions between patients with psychosis depending on presence/absence of hallucinations and/or delusions.

49 patients with DSM psychosis diagnosis (42 men and 7 women; mean age: 40), who attended a Mental Health Rehabilitation Service in 2008, of whom 7 only deluded, 6 only hallucinated, 11 deluded-hallucinated and the remaining 25 neither hallucinated nor deluded.

A Cross-sectional design (one measurement) for a co-relational method of comparison between groups.

We used the Spanish validated Anxiety Sensitivity lndex-3 -ASI 3- (Sandín et al., 2007), a 18-item Likert self-report that assesses fears of anxious symptoms. It presents a hierarchical structure (a general factor and three subscales -Physical, Cognitive and Social Concerns-). It's also used the first and third items (delusions and hallucinatory behaviour) of The Positive and Negative Syndrome Scale -PANSS- (Kay, Opler and Lindenmayer, 1988) to detect positive symptoms.

All analysis were accepted at p < .05. Patients only hallucinators showed a higher anxiety sensitivity in Social Subscale than non-hallucinative/non-delusional patients; the former presented lower punctuations in ASI-total and ASI-cognitive than patients with hallucinations and delusions. The latter showed a higher anxiety sensitivity in Cognitive Subscale than patients who only deluded.

It's hypothetized that both delusional and hallucinative activity is necessary for emergence of cognitive anxiety sensitivity.