This paper treats the problem of stochastic comparisons of two parallel systems with independent heterogeneous components having lifetimes following exponentiated Kumaraswamy-G model. The cases of same and different parent distribution functions are considered. Majorization type partial orders-based sufficient conditions in comparing the largest order statistics in terms of the usual stochastic order, reversed hazard rate order and likelihood ratio order are obtained. The likelihood ratio order among largest order statistics is established for the heterogeneous multiple-outlier exponentiated Kumaraswamy-G models. Several numerical examples are presented for illustrations as well.