Automatic speech recognition (ASR) has been significantly advanced with the\nuse of deep learning and big data. However improving robustness, including\nachieving equally good performance on diverse speakers and accents, is still a\nchallenging problem. In particular, the performance of children speech\nrecognition (CSR) still lags behind due to 1) the speech and language\ncharacteristics of children's voice are substantially different from those of\nadults and 2) sizable open dataset for children speech is still not available\nin the research community. To address these problems, we launch the Children\nSpeech Recognition Challenge (CSRC), as a flagship satellite event of IEEE SLT\n2021 workshop. The challenge will release about 400 hours of Mandarin speech\ndata for registered teams and set up two challenge tracks and provide a common\ntestbed to benchmark the CSR performance. In this paper, we introduce the\ndatasets, rules, evaluation method as well as baselines.\n