This study reports the second shared task named as UrduFake@FIRE2021 on identifying fake news detection in Urdu language. This is a binary classification problem in which the task is to classify a given news article into two classes: (i) real news, or (ii) fake news. In this shared task, 34 teams from 7 different countries (China, Egypt, Israel, India, Mexico, Pakistan, and UAE) registered to participate in the shared task, 18 teams submitted their experimental results and 11 teams submitted their technical reports. The proposed systems were based on various count-based features and used different classifiers as well as neural network architectures. The stochastic gradient descent (SGD) algorithm outperformed other classifiers and achieved 0.679 F-score.