In this article, the approximate Bayes estimation problem for the log-Dagum distribution with three parameters is considered. Firstly, the maximum likelihood estimators and asymptotic confidence intervals based on these estimators for unknown parameters of log-Dagum distribution are constructed. In addition, approximate Bayes estimators under squared error loss function for unknown parameters of this distribution are obtained using Tierney and Kadane approximation. A Monte-Carlo simulation study is performed to compare performances of maximum likelihood and approximate Bayes estimators in terms of mean square errrors and biases. Finally, real data analysis for this distribution is performed.