Our purpose was to determine which HLA class II alleles are associated with Turkish alopecia areata patients. Also we investigated whether there was a relationship between the age of onset and severity of disease and HLA alleles or not. Sixty-five patients with alopecia areata were included in this study, and 50 healthy transplant donors were used as a control group. The total group of alopecia areata patients as well as various subgroups according to scalp hair loss were compared to the control group. HLA DNA typing was performed by polymerase chain reaction/sequence specific primer method. The frequency of DQB1*03 allele was 86.1% in all patients compared to 62.0% in controls (P = 0.005). While the frequency of DQB1*03 was significantly increased, the frequency of DRB1*03 was decreased in the all patients group (4.6% versus 22.0%, P = 0.01). In the group of scalp hair loss less than 25%; the frequency of DRB1*03 was decreased (3.2%, P = 0.02). The group of patients with 25-75% scalp hair loss was compared to control group; the frequencies of DRB1*04 (66.7% versus 28.0%, P = 0.02) was increased. When the alopecia totalis, alopecia universalis or alopecia totalis/alopecia universalis group was compared to control group; DQB1*03 was associated with an increased frequency in this group versus control group (90.9%, P = 0.03). There were no significant differences for the other DQ alleles and the DR alleles tested in the patients and in the controls. When patients with early onset were compared to patients with late onset; no significant allele differences were found. Our findings suggested that DQB1*03 allele is a marker for general susceptibility to alopecia areata and may also serve as special genetic marker for susceptibility for the severe form of alopecia areata in our population. However, this association is not related to age at onset of the disease.