Lidar is a remote sensing technology that uses laser beams to generate high-accuracy, three-dimensional (3D) information of the Earth. As urban areas are developing and expanding rapidly, lidar applications such as 3D building modelling and city mapping are of increasing importance. Hence building boundary extraction is one of the main applications of lidar in civil engineering and urban planning projects. In this paper, three boundary extraction algorithms including an alpha-shape algorithm, a modified concave hull algorithm and a grid-based algorithm are tested to assess their object-by-object accuracy. The alpha-shape algorithm generates reliable boundaries for most of sample buildings, while the grid-based algorithm shows less consistency in some cases. The concave hull algorithm performs moderately with a few limitations. Advantages and disadvantages of each algorithm are identified and addressed in this paper.