By conducting three-dimensional simulation with consideration of small-strain characteristics of soil stiffness, the effects of excavation geometry and tunnel cover to diameter ratio on deformation mechanisms of an existing tunnel located either at a side of basement or directly underneath the basement were systematically studied. Field measurements were used to verify the numerical model and model parameters. For basement excavated at a side of an existing tunnel, the maximum settlement and horizontal displacement of the tunnel are always observed at the tunnel springline closer to basement and tunnel crown, respectively, regardless of basement geometry. By increasing basement length and width by five times, the maximum movements of tunnel located at the side of basement and directly underneath the basement increase by 450% and 186%, respectively. Obviously, tunnel movements are more sensitive to basement length rather than basement width. For basement excavated at a side of an existing tunnel, tunnel movements at basement centerline become stable when basement length reaches 10 He (i.e., final excavation depth). Moreover, tunnel heaves due to overlying basement excavation become stable when the normalized basement length (L/He) is larger than 8.0. As tunnel cover to diameter ratio varies from 2.5 to 3.0, the maximum heave and tensile strain of tunnel due to overlying basement excavation decrease by up to 41.0% and 44.5%, respectively. If basement length is less than 8 He, the assumption of plane strain condition of basement-tunnel interaction grossly overestimates tunnel movements, and ignores tensile strain of tunnel along its longitudinal direction. Thus, three-dimensional numerical analyses are required to obtain a reasonable estimation of tunnel responses due to adjacent and overlying basement excavations in clay.