For the past years, a better and complete 3D building model has been intensely studied. To get a 3D building model as desired, it is known to use data fusion which is a technique of 3D dataset combination. Data fusion is conducted to overcome the limitations of each measurement technique in obtaining data. This paper aims to show the status of 3D building models development that is generated from two or more datasets by using data fusion technique. From several works of data fusion, it is observed the development of data fusion related things like various technologies, methods, and accuracy. The technology used varies for data fusion, for both air and ground based measurements. Technologies such as Terrestrial Laser Scanning (TLS) and close-range photogrammetry are the most appropriate surveying methods to generate accurate and high-resolution models. Meanwhile, recent technology of mobile laser scanning makes scans more quickly and offers more convenience to obtain data in a difficult area. Furthermore, mobile laser scanning has higher completeness data than photogrammetry data. For future work, mobile laser scanning data is be considered to use in data fusion. In term of accuracy, several works agreed that data fusion is a better way to have more complete 3D building models with high accuracy. Related to data processing of data fusion, most of algorithms have disadvantage in accelerating the data fusion time. Therefore, an algorithm to shorten the processing time needs to be created.