This paper concerns self-consistent estimators for survival functions
based on doubly censored data. We establish strong uniform consistency
, asymptotic normality and asymptotic efficiency of the estimators und
er mild conditions on the distributions of the censoring variables.