The paper introduces a framework called Summarized Retrieval (SuRE) to enhance open-domain question answering (ODQA) using large language models (LLMs). SuRE addresses the limitations of existing methods that require additional fine-tuning, which is infeasible with recent LLMs. SuRE constructs summaries of retrieved passages for multiple answer candidates and evaluates their validity and ranking to select the most plausible answer. Experimental results on various ODQA benchmarks show that SuRE improves accuracy by up to 4.6% in exact match (EM) and 4.0% in F1 score compared to standard prompting approaches. SuRE is also shown to be compatible with different retrieval methods and LLMs, and its generated summaries provide additional advantages in measuring the importance of retrieved passages and serving as preferred rationales. The framework is designed to be simple and effective, making it suitable for various real-world applications.The paper introduces a framework called Summarized Retrieval (SuRE) to enhance open-domain question answering (ODQA) using large language models (LLMs). SuRE addresses the limitations of existing methods that require additional fine-tuning, which is infeasible with recent LLMs. SuRE constructs summaries of retrieved passages for multiple answer candidates and evaluates their validity and ranking to select the most plausible answer. Experimental results on various ODQA benchmarks show that SuRE improves accuracy by up to 4.6% in exact match (EM) and 4.0% in F1 score compared to standard prompting approaches. SuRE is also shown to be compatible with different retrieval methods and LLMs, and its generated summaries provide additional advantages in measuring the importance of retrieved passages and serving as preferred rationales. The framework is designed to be simple and effective, making it suitable for various real-world applications.