This paper provides a comprehensive review of vision-based 3D occupancy prediction in autonomous driving, a task that predicts the spatial occupancy status and semantic categories of 3D voxel grids around an autonomous vehicle from image inputs. The authors discuss the background, challenges, and recent advancements in this field, categorizing existing methods into three main groups: feature enhancement, deployment friendliness, and label efficiency. They analyze each category in detail, highlighting the strengths and limitations of different approaches. The paper also presents a summary of current research trends and proposes future directions, including the need for more efficient data, methodologies, and tasks. Additionally, a regularly updated collection of related papers, datasets, and codes is organized on a GitHub repository to support further research. The contributions of the paper include being the first comprehensive review of vision-based 3D occupancy prediction, providing in-depth analysis and comparisons, and offering inspiring future outlooks.This paper provides a comprehensive review of vision-based 3D occupancy prediction in autonomous driving, a task that predicts the spatial occupancy status and semantic categories of 3D voxel grids around an autonomous vehicle from image inputs. The authors discuss the background, challenges, and recent advancements in this field, categorizing existing methods into three main groups: feature enhancement, deployment friendliness, and label efficiency. They analyze each category in detail, highlighting the strengths and limitations of different approaches. The paper also presents a summary of current research trends and proposes future directions, including the need for more efficient data, methodologies, and tasks. Additionally, a regularly updated collection of related papers, datasets, and codes is organized on a GitHub repository to support further research. The contributions of the paper include being the first comprehensive review of vision-based 3D occupancy prediction, providing in-depth analysis and comparisons, and offering inspiring future outlooks.