Equivocal evidence exists regarding the degree of cognitive stability and prevalence of cognitive impairment in very late life. The objective of the current study was to examine mental status performance and change over time within a sample of Iowa centenarians. The baseline sample consisted of 152 community-dwelling and institutionalized centenarians. Twenty eight of these centenarians participated in the next phase of the study which involved up to four subsequent assessments conducted over a 6-month period. Centenarians' Short Portable Mental Status Questionnaire (SPMSQ; Pfeiffer, 1975) performance indicated that 40% of the baseline sample performed within the normal range while an additional 40% of centenarians' performed within the moderately or severely impaired range. Examination of individual-level change indicated four patterns of short-term longitudinal performance which depicted stability, enhancement, decrement, and variability in scores across the 8-month testing period. The degree of reliable change as assessed by standard error of measurement largely mirrored change as classified by the traditional scoring categories, however, the former approach appeared more sensitive to meaningful intraindividual change in later assessments.