Recommendation systems have witnessed significant advancementsand have been widely used over the past decades. However, mosttraditional recommendation methods are task-specific and therefore lack efficient generalization ability. Recently, the emergenceof ChatGPT has significantly advanced NLP tasks by enhancingthe capabilities of conversational models. Nonetheless, the application of ChatGPT in the recommendation domain has not beenthoroughly investigated. In this paper, we employ ChatGPT as ageneral-purpose recommendation model to explore its potentialfor transferring extensive linguistic and world knowledge acquiredfrom large-scale corpora to recommendation scenarios. Specifically,we design a set of prompts and evaluate ChatGPT’s performanceon five recommendation scenarios, including rating prediction,sequential recommendation, direct recommendation, explanationgeneration, and review summarization. Unlike traditional recommendation methods, we do not fine-tune ChatGPT during the entireevaluation process, relying only on the prompts themselves to convert recommendation tasks into natural language tasks. Further,we explore the use of few-shot prompting to inject interaction information that contains user potential interest to help ChatGPTbetter understand user needs and interests. Comprehensive experimental results on Amazon Beauty dataset show that ChatGPT hasachieved promising results in certain tasks and is capable of reaching the baseline level in others. We conduct human evaluationson two explainability-oriented tasks to more accurately evaluatethe quality of contents generated by different models. The humanevaluations show ChatGPT can truly understand the provided information and generate clearer and more reasonable results. Wehope that our study can inspire researchers to further explore thepotential of language models like ChatGPT to improve recommendation performance and contribute to the advancement of the recommendation systems field. The prompts and codes are availablein https://github.com/williamliujl/LLMRec.