OBJECTIVES: To investigate whether a low peripheral blood lymphocyte count is associated with increased mortality risk in older persons and to determine whether this association could be ascribed to ill health. DESIGN: A cohort study with a total follow-up period of 1,602 person years. SETTING: Leiden, the Netherlands. PARTICIPANTS: Four hundred thirty-six community-dwelling residents aged 85 and older. MEASUREMENTS: Health status and leukocyte total and differential counts were assessed at baseline. Lymphocyte subsets were measured with a fluorescence-activated cell sorter. Age- and sex-adjusted mortality risks were estimated using Cox proportional hazard regression analysis. RESULTS: There was no association between lymphocyte count and mortality in persons with ill health (mortality risk lowest vs highest quartile = 1.16; 95% confidence interval (CI) = 0.85-1.58, P = .35), but mortality was dependent on lymphocyte count if disease was excluded (mortality risk lowest vs highest quartile = 2.14; 95% CI = 1.08-4.23, P = .03). A similar increase in mortality risk was found when the cluster designation (CD)4(+), CD8(+), and CD16(+) lymphocyte subsets were analyzed. Within individuals, low values of the lymphocyte subsets were related and there was no compensatory increase in CD16(+) lymphocyte counts. A low lymphocyte count was not associated with specific causes of death. CONCLUSION: A low lymphocyte count was associated with an increased mortality risk in older persons without apparent disease. This association was not only found for the total lymphocyte count but also for the CD4(+), CD8(+), and CD16(+) lymphocyte subset counts.