The Attitudes and Beliefs Scale (ABS) is a widely used measure of irrational beliefs (IBs) but has important psychometric problems. Our objective is to improve the psychometric quality of a Spanish version of the scale. Classical test theory, item response theory, and confirmatory factor analyses were combined to obtain a shorter version of the scale using 2 samples: one from the general population (n= 565) and another with chronic pain (n= 514). Pearson correlations were performed with IBs, personality and health measures to investigate sources of construct validity. After eliminating half of the items (12), the factorial fit of the scale became very good (RMSEA< .08; CFI and TLI> .95). IBs were associated with more neuroticism (.21≤ r≤ .61, p≤ .001) and poorer mental health (-.17≤ r≤ -.56, p≤ .001), as well as a less extraversion and conscientiousness (-.14≤ r≤ -.41, p≤ .01). These results were replicated in both samples, but IBs were only associated with poorer physical health in the general population sample. The shortened Spanish version of the ABS is a valid and reliable instrument that can be rapidly administered in clinical settings.