On the basis of prior research (L6pez & Romero, 1988), we examined whether using the Wechsler Adult Intelligence Scale-Revised (WAIS-R) can lead to an underestimation of Spanish-speaking older adults' functioning and whether using the Escala de Inteligencia Wechsler para Adultos (E1WA) can lead to an overestimation of this group's functioning. Spanish-speaking patients who met criteria for Alzheimer's Disease (AD) and Spanish-speaking nonimpaired elderly were tested with selected subtests from the WAIS-R and EIWA, and the Mini-Mental State Examination. The AD patients' activities of daily living were also assessed. For purposes of comparison, Englishspeaking nonimpaired older adults and AD patients were administered the same tests except that the Wechsler subtests were all taken from the WAIS-R. The results point out clear differences between impaired and nonimpaired groups, and differences and similarities between the two language groups. We interpret the differences between the language groups as reflecting test bias when using either the WAIS-R or EIWA in assessing Spanish-speaking elderly. The findings have implications for defining culturally sensitive psychological assessment. The psychologist who chooses to assess Spanish-speaking individuals is confronted with a major challenge: to conduct psychological assessment in a culturally sensitive fashion. Lopez et al. (1989) denned cultural sensitivity as carefully balancing general norms with culture-specific norms in judging a specific patient or client. Consider the example of an elderly women with limited knowledge of English who performs poorly on an English language test such as the Wechsler Adult Intelligence Scale-Revised (WAIS-R; Wechsler, 1981). In adopting an etic perspective, a perspective that applies what are thought to be general norms, one may then interpret the test results as reflecting impairment. Compared with the English language standardization sample, this predominantly Spanish-speaking elderly women is functioning at the same level as English-speaking individuals who are impaired in their cognitive-intellectual functioning. In this case, however, it is possible that the belowaverage performance may reflect limited English language skills, an unfamiliarity with the American culture, a limited educational background, or all of these factors. Therefore, when applying etic norms, in this case WAIS-R norms, it seems possible that psychologists could underestimate the cognitiveintellectual functioning of Spanish-speaking adults.