This study explores the way translation crowdsourcing may contribute to metacognitive translator training through a teaching experiment where Global Voices Lingua was integrated into an undergraduate English-Chinese translation course. In doing so, the study investigates how translation students' awareness of conditional knowledge and its monitoring functions may be raised. Each of the 15 students enrolled in the translation course was asked to translate a Global Voices' English news post into Chinese and record the translation process. Then, the students were instructed to write self-reflections on their problem-solving and decision-making processes based on their recording files. The results reveal that the students solved approximately 73% of the 215 translation problems encountered by considering relevant situational demands explicitly or implicitly. Further, the monitoring functions recognised by the students include: (1) avoiding literal translation, (2) improving readability, and (3) choosing an appropriate translation. Among the 200 cases where the monitoring functions were at play, only 11.5% were misapplied or unjustifiable. Overall, this study's findings indicate that translation crowdsourcing can offer reliable authentic practices, strengthen students' metacognitive abilities, and help them produce translations that fulfil contextual demands.