HLA DR4 antigens have been considered as a risk factor in periodontal disease. The aim of this "case control" study was to verify and to provide fuller clarification of such data. "Cases" or patients had to be aged between 20 and 48 years. They presented at least 5 sites spread over several teeth with an attachment loss equal or greater than 6 mm, and 10 sites spread over several teeth with periodontal pockets equal to or greater than 5 mm. Verification with a WHO probe showed an individual CPITN score of 4. Moreover, subjects whose average CPITN score for the 6 sextants was less than 3 were excluded from the study. Among these "severe periodontitis" patients, a subgroup was distinguished composed of subjects aged 20-35 years who presented, in accordance with the cases by Katz and co-workers, 5 or more teeth showing pocket depths of 6 mm or more. The dental chartings of these subjects showed an attachment loss of more than 3 mm on certain teeth over an inter-exam period of 1-3 years. They all displayed obvious loss of bony support in the affected sites. This constituted the "rapidly progressive periodontitis" subgroup. The "controls" were all over 20 years of age, and it was clinically verified that they were free of periodontal disease. There were 48 "cases" and 55 "controls". HLA typing of patients and controls was performed using "sequence oligoprobe hybridization after polymerase chain reaction" in accordance with the 11th International Workshop. This method allowed the detection of DR4 alleles as well as DR4 subtypes. The ethno-geographic origin of the subjects, considered as a confounding variable, was neutralized by stratified analysis. Subtypes 0401, 0404, 0405 and 0408 tended to be more frequent (p=0.08) in the cases (Severe Periodontitis). Focusing on analysis of "rapidly progressive periodontitis" in subjects aged 20-35 years, a very significant Mantel-Haenszel chi2 was obtained (p=0.0058) which led to a Mantel-Haenszel standardized odds ratio (OR) equal to 17. The 95% confidence interval was 1.03<0.R.<180.10. In conclusion, this supports previous reports and gives further clarification: in particular subtypes 0401, 0404, 0405 and 0408 can be considered as a risk factor for "rapidly progressive periodontitis". It should be noted that these determinants have been implicated in rheumatoid arthritis.