Abstract Since X-ray diffraction patterns are directly related to crystal structures. X-ray identification is, in principal, better suited to the recognition of structural groups and structural varieties than of chemical species. Well-formed kaolin, mica, and chlorite structures give rise to characteristic 7, 10 and 14A spacings which are relatively easily identified. Hydrated forms, such as hydrated halloysite (d=10A) and montmorilionoids under normal conditions (d=14A), are recognized either by low-temperature dehydration giving characteristically diminished spacings, or by the formation of organic complexes giving characteristically increased spacings. This introduces at once the principle that X-ray identification may, and generally does, entail the study of mineral modification by suitable chemical and/or thermal treatment.The main requirements in the X-ray technique are: (1) Ability to record long spacings up to 25A or even higher values; (2) well focused lines with good resolution; (3) absence of background scattering and white radiation anomalies. Of these, (1) and (3) lie within the control of the investigator, but (2) depends partly on the material.The main difficulties of X-ray identification are: (1) The multiplicity of lines when many components are present, and (2) poorly defined diagrams arising from poor crystallinity and/or small size of crystals. The first can be treated experimentally, but the second is inherent in the problem. Simplification of X-ray diagrams is possible by sedimentation, acid treatment and thermal treatment and by the use of orientated specimens. Poor quality powder diagrams (if not due to poor technique) are of interest in themselves and may still give valuable information concerning the clay minerals, more particularly when monomineralic or mainly monomineralic fractions are used. Line profiles give valuable indications of randomly displaced layers and of interstratified sequences.The recognition of chemical species by X-rays is difficult but some progress can be made if pure or nearly pure specimens are obtainable. The recognition of di- and tri-octahedral minerals by the (060) spacing is well known but the recognition of micas and chlorites in any greater detail is difficult. A consideration of basal spacings and basal intensities may make more specific identification possible.