The experimental results which are obtained from electron nuclear double resonance (ENDOR) of point defects in solids are briefly reviewed. From the relation between the expe¡ superhypeffine and quadrupole interaction tensors and the spin density distribution and charge density dist¡ the detalls of the microscopic structure and thus the positions of atoms can in principle be obtained. For this, in general, a precise theory of the defect wave function is required. However, there are many cases in which simple approximations to the wave function can be used to determine atomic positions rather well. This is illustrated with several examples, mainly from ionic solids. The particular difficulties encountered in semiconductors and the influence of dynamical effects are also discussed.