Este artículo tiene como objetivos estudiar la estructura factorial, consistencia interna, distribución, diferencias entre sexos, validez convergente y relación con la edad y deseabilidad social del Inventario de ansiedad y fobia social (Social Phobia and Anxiety Inventory [SPAI]). En una muestra de 695 estudiantes universitarios con cuotas equivalentes por sexos se aplicaron el SPAI, la Escala de ansiedad en la interacción social (Social Interaction Anxiety Scale [SIAS]), la Escala de autoverbalizaciones al hablar en público (Self-Statements during Public Speaking Scale [SSPS] y el Inventario balanceado de deseabilidad social al responder (Balanced Inventory of Desirable Responding [BIDR]). Reducido el SPAI a 41 ítems, el modelo estructural de un factor de fobia social (con cinco factores subordinados) correlacionado con un factor de agorafobia mostró un ajuste adecuado a los datos. Los valores de consistencia interna fueron altos, las distribuciones asimétricas positivas, las mujeres promediaron más alto que los hombres, las correlaciones fueron bajas con edad y el factor de manejo de la impresión del BIDR, y moderadas con SIAS y SSPS. Se sugiere el uso del SPAI-41 en México y su estudio en otros países hispanoparlantes. The aims of this paper were to study the factorial structure, internal consistency, distribution, sex differences, convergent validity, and the relation to age and social desirability of the Social Phobia and Anxiety Inventory (SPAI). The SPAI, Social Interaction Anxiety Scale (SIAS), Self-statements Scale during Public Speaking (SSPS) and Balanced Inventory of Desirable Responding (BIDR), were administered to a sample of 695 college students with equivalent gender quotas. If the SPAI is reduced to 41 items, a structural model of one factor of social phobia (with five subordinate factors) correlated with one factor of agoraphobia showed an adequate fit to the data. The internal consistency values were high, distributions were positively skewed, women averaged higher than men, and correlations were low with age and the BIDR impression management factor. They were moderate with the SIAS and SSPS. We suggest the use of the SPAI-41 in Mexico and its study in other Spanish-speaking countries