HLA class II genes were analyzed to study IDDM susceptibility in Cantabria (Northern Spain). Patients showed highly significant increases in DRB1*0301 (RR = 4.581, p < 0.00005), DRB1*0401 (RR = 2.6, p < 0.05), DRB1*0402 (RR = 8.78, p < 0.05) and DRB1*0405 (RR = 14.73,p < 0.005). Highly significant differences were in the DQA1*0301 (RR = 3.62, p < 0.000005) and DQA1*0501 (RR = 2.13, p < 0.05) alleles. DQB1*0201 (RR = 4.1, p < 0.00005) and DQB1*0302 (RR = 5.42, p < 0.000005) alleles were also significantly increased. A significant increase in DRB1*0402-DQA1*0301-DQB1*0302 (RR = 16.18, p < 0.05), DRB1*0405-DQA1"0301-DQB1*0302 (RR = 16.12, p < 0.05), DRB1*0301-DQA1*0501 DQB1*0201 (RR = 4.58, p < 0.00005) and DRB1*0401-DQA1*0301-DQB1*0302 (RR = 4.36, p < 0.005) was apparent in the diabetic group, while the DRB1*1501- DQA1*0102-DQB1*0602 and DRB1*1401-DQA1 *0104-DQB1*05031 protective haplotypes (RR = 0.17 and 0.09, p < 0.0005 and 0.05, respectively) were significantly lower in patients. The absence of Asp57 and the presence of Arg52 were associated with disease in a dose-dependent: manner. Several genotypes encoding the identical DQ alpha 52/DQ beta 57 phenotype carried very different RRs. Finally, the Cantabrian population has the highest incidence of IDDM reported for Spain (15.2 of 100.000 in the 0-14 age group, Poisson's 95% CI: 10.6-19.3). (C) American Society for Histocompatibility and Immunogenetics, 1999. Published by Elsevier Science Inc.