This study investigates the impact of machine translation on the cognitive decision-making processes of translators from the perspective of information searching. The study has compared and analyzed the information searching process of machine translation post-editing and human translation. Key-logging tool Translog II and screen recording of BB FlashBack Pro 5 were used, supplemented with pre- and post-test questionnaires. The study focused on three dimensions: time allocation during processing, information search patterns, and translation quality. The findings show that: (1) machine translation post-editing saves translators’ time on information search but requires more time for the planning stage; (2) the number of information-searching instances in post-editing was lower than those in human translation, with search behaviors primarily focused on semantic retrieval of vocabulary; (3) the quality of post-editing quality is better than that of manual translation. Translation learners hold a positive attitude towards the use of machine translation, as machine translation helps in resolving vocabulary searching. The findings provide empirical support and useful reference points for optimizing information search in translation and advancing the reform of machine translation post-editing.