[1] He played a role in the early development of input-output analysis and linear programming.
Cornfield played a crucial role in establishing the causal link between smoking and incidence of lung cancer.
[2] He introduced the Rare disease assumption[3] and the "Cornfield condition" that allows one to assess whether an unmeasured (binary) confounder can explain away the observed relative risk due to some exposure like smoking.
He later worked at the National Cancer Institute, the Department of Biostatistics at Johns Hopkins School of Hygiene and Public Health, the National Heart Institute, the University of Pittsburgh, and George Washington University.
[5] He was the R. A. Fisher Lecturer in 1973 and President of the American Statistical Association in 1974.