In a prospective study the degree of hip joint laxity was compared with subsequent hip joint development. A total of 70 non-selected adult dogs from 34 breeds presented for hip dysplasia screening was examined. The degree of hip joint laxity was quantitated using a newly developed radiographic stress technique. The dogs were re-examined according to the standard radiographic technique after 1 year or more. There was a positive correlation (r = 0.58) between the degree of subluxation (SI) and the subsequent Canine hip dysplasia (CHD) grade. All dogs with an SI value of up to 0.35 developed hip joints ranging between normal and mildly dysplastic at worst. Of the dogs with an SI value higher than 0.35, 76% developed hip dysplasia. Of those 40 dogs grading normal or borderline normal on standard radiographs, 15% demonstrated lax hip joints with an SI of more than 0.35. Breed-specific differences in the correlation of degree of hip joint laxity to the development of coxarthrosis were noted. If both excessive hip joint laxity and development of arthrosis are considered exclusion criteria for breeding dogs, then the current mode of selection does not adequately restrict potential breeding stock. Only dogs demonstrating an SI of 0.35 or less on stress radiographs and graded normal or borderline normal on standard radiographs should be used for breeding, equivalent to 49% of all dogs examined in this study. A further 36% of the dogs examined, most of them currently still accepted as breeding dogs, would not pass. Breed-specific acceptable degree of hip joint laxity in breeding dogs should be determined. Selecting breeding dogs based on the results of hip joint laxity assessment may further decrease the incidence of CHD in the offspring.