In this thesis, we develop inference procedures for copula-based models of bivariate dependence. We first investigate the distribution of Kendall’s functions for joint survivors since Kendall’s functions
are important for identifying Archimedean copula models. We then provide two estimators for the generator of an Archimedean copula. We also propose a plug-in estimator for Kendall’s tau and a
maximum pseudo-likelihood estimation for the copula model parameters in cases where the survival times are subject to bivariate random censoring. These tests are based on the usual (univariate) Kaplan-Meier estimator and a recently proposed estimator for the bivariate case.