The random errors inherent in two common types of x-ray diffractometer measurement have been estimated. Most of the error is apparently associated with operations contingent upon setting up the specimen and getting the instrument in motion. Although significant mean differences were detected between mounts of different type, different mounts of the same type gave agreement well within experimental error. Results obtained at scale factor 4 were indistinguishable from those obtained at scale factor 2. Scanning direction exerts a small but significant effect on estimates of distances between peaks. The exact numerical results apply strictly only to the specific operator-instrument combination which generated the data. The evaluation procedures should be widely applicable.