This paper provides a comprehensive review of evaluation methods for large language models (LLMs), focusing on three key dimensions: what to evaluate, where to evaluate, and how to evaluate. The authors, from various institutions in China and the USA, highlight the importance of LLMs in both academic and industrial settings due to their exceptional performance in various applications. The paper covers a wide range of evaluation tasks, including natural language processing, reasoning, medical usage, ethics, education, natural and social sciences, and agent applications. It also discusses the selection of appropriate datasets and benchmarks for evaluation and explores current evaluation protocols and novel approaches. The authors summarize the success and failure cases of LLMs in different tasks, emphasizing the need for more in-depth research and optimization. Additionally, the paper addresses future challenges in LLMs evaluation, such as ensuring safety and reliability in critical sectors and designing new evaluation protocols. The authors aim to offer valuable insights to researchers in the field of LLMs evaluation and foster a collaborative community for better evaluations.This paper provides a comprehensive review of evaluation methods for large language models (LLMs), focusing on three key dimensions: what to evaluate, where to evaluate, and how to evaluate. The authors, from various institutions in China and the USA, highlight the importance of LLMs in both academic and industrial settings due to their exceptional performance in various applications. The paper covers a wide range of evaluation tasks, including natural language processing, reasoning, medical usage, ethics, education, natural and social sciences, and agent applications. It also discusses the selection of appropriate datasets and benchmarks for evaluation and explores current evaluation protocols and novel approaches. The authors summarize the success and failure cases of LLMs in different tasks, emphasizing the need for more in-depth research and optimization. Additionally, the paper addresses future challenges in LLMs evaluation, such as ensuring safety and reliability in critical sectors and designing new evaluation protocols. The authors aim to offer valuable insights to researchers in the field of LLMs evaluation and foster a collaborative community for better evaluations.