Objectives.To assess psychometric properties and to develop a shorter version of a new anxiety scale. Materials and methods: The 130-item new anxiety scale (NAS-130) was administered to 254 psychiatric outpatients of a general hospital in Lima, Peru. A categorical principal component analysis (CATPCA) was performed with the 130 items and based on these results and the relation of each item with a Clinical Global Impressions Severity Scale for anxiety (CGI-S) the scale was simplified. In addition, the CGI-S was used to assess the concurrent validity of the scale. Results. The NAS-130 had a Cronbach's alpha of 0.97 and a Spearman correlation coefficient (Rho) with CGI-S of 0.44 (p<0.01). After item selection, a 72-item scale was obtained (NAS-72) which had a Cronbach's alpha of 0.96 and a Rho with CGI-S of 0.47 (p<0.01). All items of NAS-72 had saturation coefficients greater than 0.4 in one dimension according to the CATPCA. NAS-130 and NAS-72 scores were significantly higher in women than in men (p<0.01). Conclusions. The NAS-72 has psychometric characteristics that suggest that it could be useful to measure anxiety in our population; however, the scores should be interpreted differently in men and women.