Francesco Cucca and colleagues (Cell 155, 242–256, 2013) have applied a sequencing-based genome-wide association approach to analyze how genetic variants influence the relative abundance of immune cell types. The authors collected peripheral blood samples from 1,629 individuals from Sardinia and used fluorescence-activated cell sorting to purify 95 distinct immune cell types, with an emphasis on specialized T cell subsets. They then tested more than 8 million genetic variants for association with 272 different traits related to immune cell levels or function, with further replication testing in 1,241 additional samples. In total, they observed 23 genome-wide significant association signals at 13 loci, 3 of which were previously associated with altered disease risk. For example, in the IL2RA region, they found that a variant previously associated with reduced susceptibility to type 1 diabetes was associated with higher levels of a memory T cell subset characterized by high CD25 expression, suggesting that this cell type might confer protection against this autoimmune disease. Further exploration of coincident associations between disease risk and immune cell phenotypes could provide new insights into disease mechanisms or suggest possible strategies for therapeutic intervention.