||There have been an increasing number of tunnels constructed over the past decades due to rapid development in urban areas. It is sometimes inevitable whereby multiple tunnels are constructed in a closely spaced area in order to develop more efficient and environmentally friendly infrastructure in congested cities. Tunnels may be parallel (side-by-side) or vertically aligned (piggyback tunnelling). Studies on multiple tunnel interaction are getting popular these days as the impact on the existing tunnels due to newly constructed tunnels may be tremendous in terms of serviceability and safety problems. However, the soil-structure interaction problems arising from perpendicularly crossing tunnels attract relatively little research attention. In the past, limited studies related to twin tunnel interaction have been conducted by considering two-dimensional tunnel excavation and simulating the effects of volume loss only. Subsequently, effects of stress transfer caused by multi-stage tunnel advancement induced by different construction sequences and weight loss on an existing tunnel with different C/D ratios are not fully understood. In view of the afore mentioned issues, this research aims at investigating the effects of (a) weight and volume losses (b) different construction sequences and (c) different C/D ratios on perpendicularly crossing tunnels. In this research, a series of three-dimensional centrifuge model tests are conducted to investigate the effects of weight and volume losses on the interaction between perpendicularly crossing tunnels. To simulate a three-dimensional tunnel advancement process with the effects of both weight and volume losses, a novel and newly developed technique is adopted. The technique comprises of six independent pairs of volume-controllable rubber bags containing heavy fluid. An outer rubber bag is used to control volume loss, whereas an inner rubber bag is adopted to simulate weight loss. In addition, numerical analysis is carried out to back-analyse the centrifuge test results and explain the stress transfer mechanism due to twin tunnel interaction. Besides that, additional numerical analysis is carried out to study the influence of weight loss for tunnelling with different construction sequence. The maximum measured induced settlement for case E3N2 (i.e. new tunnel constructed above existing tunnel) is two times larger than that induced in case E2N3 (i.e. new tunnel constructed below existing tunnel) and case E5N3 (i.e. new tunnel constructed above existing tunnel). The reason is not only the C/D ratio of new tunnel in case E3N2 is smaller but also the presence of the existing tunnel stiffens the soil above the new tunnel for case E2N3 and widens the settlement trough. The existing tunnel experience vertical elongation and horizontal compression due to vertical stress release during tunnelling for all three cases. Among all three cases, the normal stress reduction for cases E2N3 is the largest. As a result, the displacement and deformation of existing tunnel and the transverse bending moment induced in lining of the existing tunnel for case E2N3 is larger than cases E3N2 and E5N3. Also, the influence zone due to the effect of stress release is within a distance of 1D from the existing tunnel axis for all three cases. The maximum ground surface settlement for cases considering only effects of volume loss without effects of weight loss are smaller than those considering both weight and volume losses. This is because the effect of weight loss results in additional stress release around the new tunnel which increases the ground surface settlement. Besides that, the transverse bending moment for the case only considering volume loss is 54% smaller than the case considering both weight and volume losses. The is due to the difference in normal stress i.e. stress release acting on the existing tunnel by 62%.