The methodological shortcomings characterizing previous experimental research on selective and cross-cutting exposure (e.g. focusing on the impact of a few experimental factors and providing a limited range of choice alternatives) amount to an incomplete description of news choice behavior. This study proposes an alternative method that addresses these shortcomings in testing partisan news selection in the current media environment. A conjoint experiment (N = 746), which simulates a popular mobile news platform (iOS news app), showed that while partisans preferred selecting news from pro-attitudinal sources, they also selected news from counter-attitudinal sources at least once in five times (20%). The tendency to engage in selective exposure was slightly stronger among strong partisans. Our findings indicate that the concerns over partisan selective exposure may not be an issue of “selection” but more of “access” to news sources with different political views. The implications for future research are discussed.