The paper introduces a novel method called LLM Assisted Causal Recovery (LACR) that leverages large language models (LLMs) to recover causal graphs. LACR combines retrieval-augmented generation (RAG) to enhance the LLMs' knowledge base with scientific publications and experimental data, and a prompting strategy to extract associational relationships among variables. The method aims to address the limitations of traditional causal graph recovery methods, which often suffer from data collection biases and individual knowledge constraints. LACR is evaluated on several benchmark datasets and shows superior performance in causal graph quality compared to other LLM-based methods. Additionally, LACR demonstrates sensitivity to new evidence in the literature, making it useful for updating causal graphs. The paper also discusses the contributions, background, methodology, experiments, and limitations of the proposed method.The paper introduces a novel method called LLM Assisted Causal Recovery (LACR) that leverages large language models (LLMs) to recover causal graphs. LACR combines retrieval-augmented generation (RAG) to enhance the LLMs' knowledge base with scientific publications and experimental data, and a prompting strategy to extract associational relationships among variables. The method aims to address the limitations of traditional causal graph recovery methods, which often suffer from data collection biases and individual knowledge constraints. LACR is evaluated on several benchmark datasets and shows superior performance in causal graph quality compared to other LLM-based methods. Additionally, LACR demonstrates sensitivity to new evidence in the literature, making it useful for updating causal graphs. The paper also discusses the contributions, background, methodology, experiments, and limitations of the proposed method.