AbstractThe problems of estimation and hypotheses testing on the parameters of two correlated linear models are discussed. Such models are known to have direct applications in epidemiologic research, particularly in the field of family studies. When the data are unbalanced, the maximum‐likelihood estimation of the parameters is achieved by adopting a fairly simple numerical algorithm. The asymptotic variances and covariances of the estimators are derived, and the procedures are illustrated on arterial‐blood‐pressure data from the liter