Virtual reality (VR) is increasingly influencing how researchers study human behavior. Yet, running large-scale VR experiments remains a challenge. The software is complex, capturing data from many sources at once is difficult, and analysis at scale takes significant effort. This paper reviews 67 open-source, research-oriented toolkits that support some or all stages of running VR studies, including the planning, design and management of experiments, collecting and synchronizing data, analyzing and visualizing results, and even running studies remotely or across different types of devices. We group existing peer-reviewed tools into five categories and highlight trends in the field. We discuss Unity's strong influence, the ongoing balance between easy-to-use interfaces and flexible coding approaches, the increased adoption of Lab Streaming Layer (LSL) for time-synchronized data capture, and growing interest in real-world and AI-supported analysis. Based on this overview, we offer practical recommendations to make VR research more consistent and sustainable, such as using shared formats for data and events, ensuring accurate timing and replay, and developing data exports that remain usable over time. We also compile a maintenance snapshot, which highlights technological and operational risks. We conclude with a roadmap targeting reproducibility, including considerations for a cross-platform VR experiment description language and distributed, replayable datasets. This review aims to help researchers choose and combine tools more effectively and give developers directions for supporting a more robust and reliable VR research ecosystem.