Importance: Emerging evidence implicates a role for neuroinflammation in Alzheimer disease (AD) pathogenesis, predominantly involving the innate immune system. Blood leukocyte counts are easily accessible markers of immune function; however, their association with the risk of AD is unknown. Objective: To investigate the observational and genetic associations between types of blood leukocytes and risk of AD. Design, Setting, and Participants: In a cohort study comprising observational and genetic analyses, the Copenhagen General Population Study prospective cohort (n = 101582) was used for the observational analyses. For the genetic studies, nonlinearity was first evaluated for the association between leukocyte cell counts and AD risk using individual-level data from the UK Biobank (n = 365913). Subsequently, a 2-sample mendelian randomization framework was applied using genetic instruments for blood leukocyte counts (n = 563 085); for AD, the European Alzheimer & Dementia Biobank was used, including 85934 individuals with AD and 401577 controls and the International Genomics of Alzheimer's Project, including 21982 individuals with AD and 41944 controls. Exposures: Observational and genetically determined types of blood leukocyte counts. Main Outcomes and Measures: Hazard ratios (HRs) and 95% CIs for AD of cell count percentile groups in observational studies and odds ratios (ORs) and 95% CIs for AD per 1 SD genetically determined cell counts. Results: This cohort study included 101582 participants (55 891 [55.0%] women) with a median age of 58 years (IQR, 48-67 years); of these, 1588 individuals developed AD. Multivariable-adjusted HRs for participants in the less than 5th vs the 25th to 75th (reference) percentile group were 1.24 (95% CI, 0.99-1.54) for blood monocytes and 1.25 for blood eosinophils (95% CI, 1.05-1.50). For participants in the greater than 95th vs the 25th to 75th percentile group, the HR was 1.30 (95% CI, 1.06-1.61) for blood neutrophils. Genetically, no evidence favored possible nonlinear associations. The ORs for AD per 1-SD decrease in genetically determined blood monocytes were 1.04 (95% CI, 1.00-1.10) in the European Alzheimer & Dementia Biobank consortium and 1.09 (95% CI, 1.01-1.17) in the International Genomics of Alzheimer's Project consortium. Using mendelian randomization, sensitivity analyses and multivariable analysis showed similar results. Conclusions and Relevance: The findings of this study suggest that low blood monocyte counts are associated with increased AD risk. These findings highlight a potential role of the innate immune system in AD pathogenesis..