Severe congenital neutropenia (SCN) was originally described as an autosomal recessive disorder. Autosomal dominant and sporadic forms of the disease have subsequently been recognized. All forms of the disease are manifest by persistent severe neutropenia and recurrent bacterial infection. Cyclical neutropenia (CyN) is characterized by periodic neutropenia inter-spaced with (near) normal neutrophil counts. Recently, heterozygous mutations in the ELA2 gene encoding neutrophil elastase (NE) have been described in the majority of cases of CyN and sporadic and autosomal dominant SCN. A case of paternal mosaicism has provided genetic "proof" of the pathogenicity of such mutations, but the exact pathogenic mechanism remains elusive. This review will focus on the mosaic proof and examine possible pathogenic mechanisms. The lack of obvious associations and indeed overlap between the mutations that cause the two diseases will also be discussed. Clinically to date, the discovery of an elastase mutation has been of limited value to individual patients. However, it is hoped that further genotype/phenotype studies may improve assessment of patient prognosis.