Double crystal x-ray topography using a synchrotron radiation source has been used to measure the lattice distortions present in 50mm diameter samples of (001) semi-insulating LEG gallium arsenide. Lattice strains and tilts have been mapped in In-doped and undoped samples as well as annealed and unannealed samples taken from the seed and tail ends of boules. The properties of the x-ray source which are necessary for these measurements are discussed and it is concluded that a synchrotron source is the only practical choice. Lattice strains of 90ppm and tilts greater than 100 arc seconds were measured in In-doped material both of which appear to be due to a combination of In concentration variations and the inhomogeneous dislocation distribution. Undoped samples were found to be more uniform with lattice strains of typically +20ppm towards the samples edges where the dislocation density is largest. The lattice tilt distribution in seed and undoped samples invariably exhibited a four-fold symmetry which was enhanced by the presence of lineage features lying along the <110> directions. Tail end samples were generally less uniform in lattice strain and showed a lower symmetry in their lattice tilts. These results are discussed in the light of current ideas concerning the origin of variations in lattice strain and EL2 concentration. An x-ray diffraction method involving integrated intensity measurements of the quasi-forbidden 200 reflection, which is highly stoichiometry sensitive, is investigated. The results, however, show no conclusive stoichiometry variations but do highlight important experimental conditions which must be satisfied if such measurements are to be meaningful. The images of dislocations in double crystal x-ray topographs are investigated and compared with theoretical simulations in order to assess the effects of point defect environment on the dislocation strain field. The results suggest that the EL2-dislocation interaction is not significantly strain driven.