Introduction: The aim of the study was to investigate a diagnostic protocol for patellar luxation (PL) in respect to its usability as a screening method in the framework of breeding programs. Further, the influence of breed, age, body mass, gender and neutering on the prevalence of PL has been investigated. Methods: In a period of 8 years (1996-2004) 432 small and miniature-breed dogs have been examined for patellar luxation. In order to achieve the diagnostic efficiency required for genetic screenings performed in the framework of breeding programs, this examination was based on the concept of a standardized examination protocol that included clinically examination with inspection as well as palpation. The diagnostic criteria were lameness, palpation and evaluation of patellar tracking in standing and recumbent position of the dog with special focus on the presence of any patellofemoral instability. A further diagnostic criterion was the deviation of the tibial tuberosity and any perceivable crepitation of the knee joint during all manipulations performed. The findings of patellar luxation were valuated according to PUTNAM's (1968) graduation. X-rays have not been performed. To find out whether the investigated diagnostic criteria fulfil the demands on consistency and validity needed for screening, diagnostic rank correlation coefficients between the single diagnostic criteria and the final PL-finding have been calculated. The influence of breed, age, body mass, gender and neutering on the occurrence of PL was investigated by calculating odds ratios using a multifactorial logistic regression model. The significance testing of the resulting odds ratios was performed by calculating the corresponding 95 % confidence intervall. Results: In 61.6 % of the examined dogs we found patellar luxation, but only in 15.5 % (right knee) and 12.8 % (left knee) a permanent lameness could be observed; in 3.5 % (right kee) and 4.6 % (left knee) there was an intermittent lameness. Therefore nearly 40 % of the animals with patellar luxation were clinically normal and would not have been detected as carriers of patellar luxation without screening diagnostic. The investigated diagnostic criteria showed significant correlation among each other and with the final PL-finding and therefore proved to be consistent and meaningful with regard to PL- investigation. The parameters "luxation in standing" and "luxation in recumbent position" showed the highest rank-correlation to the final findings. Thus it appears that, in general, this examination protocol is suitable for PL-screenings. The investigation of the influence of the parameters "body mass", "age", "gender" and "neutering" on the occurrence of PL showed that except "gender" all attributes are associated with occurrence of PL. An increase in body mass of 1 kg decreases the odds of suffering from PL to the 0.8fold (p < 0.05), an increase of age of one year increases the odds of suffering from PL to the 1.1fold (p = 0.051). Neutered dogs showed a 3.10fold odds for suffering from PL (p < 0.05). To detect breed predispositions for patellar luxation odds ratios were calculated for each breed that was represented in the study by more than 10 animals, including Jack Russell Terrier, Mops, Papillon, Pekingese, Shih Tzu, Tibet Terrier, West Highland White Terrier, Poodle, Yorkshire Terrier, Malteser and Chihuahua. 2 breeds showed a significantly different outcome with regard to increase of suffering from PL. The Jack Russell Terrier had an odds ratio of 0.31 with corresponding 95 % confidence intervals 0.14 -0.67 and therefore showed a reduced chance of PL compared with all other breeds. On the other hand the Poodles (including the miniature- and toy-variants) with an odds ratio of 5.62 with corresponding 95 % confidence intervals 1.93-16.41 showed a higher chance of PL compared with all other breeds. This result gives some evidence on a genetic background of PL. Nevertheless the genetic basis of PL should be investigated more accurately by family analyses and heritability studies. Conclusions: Referring to the consistency and validity of the diagnostic criteria the diagnostic protocol used in this study seems to be suitable as screening method for PL. As 2 of the investigated breeds showed a significantly higher or lower chance for suffering from PL a genetic background of PL can be postulated. Based on the results of this study it would be highly recommended for breeding associations to introduce screening programs for patellar luxation in small and miniature breed dogs as described in this study in combination with a redefinition of breeding goals aimed at the extreme miniaturization of the affected dog breeds.