The MICA gene encodes for major histocompatibility complex class I chain-related proteins (MIC), which belong to a recently identified new family of nonclassical major histocompatibility complex molecules. The general structure of the MICA molecule resembles that of major histocompatibility complex class I molecules. MIC molecules are considered to be stress-induced antigens that are recognized by cytotoxic T cells and natural killer cells, which play an important role in the surveillance of transformed infected and damaged cells. Associations of major histocompatibility complex class I molecules with skin cancer have been described before. To evaluate the possible association of MICA gene polymorphism with the risk for nonmelanoma skin cancer we evaluated 153 cases with squamous cell carcinoma, 261 cases with basal cell carcinoma, 111 controls with malignant melanoma, and 247 controls without a history of skin cancer. Five distinct MICA alleles A4, A5, A6, A9, and A5.1 were studied. As the MICA 5.1 variant gene contains a four-nucleotide insertion that causes a stop codon in the trans membrane region, the resulting truncated MICA molecule does not reside on the cellular membrane. In the case of individuals who are homozygous for MICA 5.1 this results in cells that are naked for the MICA molecule. We therefore specifically addressed the possible association between MICA 5.1 homozygosity and skin cancer, as these individuals are expected to be at the highest risk for skin cancer if the MICA gene plays a role in skin carcinogenesis. Viral proteins may serve as antigens for recognition of skin cancer by the immune system. Human papillomavirus is the most likely candidate virus to be involved in the carcinogenesis of cutaneous squamous cell carcinoma. Hence, we also assessed the association between MICA polymorphism and squamous cell carcinoma in human-papillomavirus-positive and human-papillomavirus-negative individuals as identified by the presence of human papillomavirus DNA in hairs plucked from their eyebrows. Our analyses did not reveal any significant differences regarding the MICA allele frequencies between cases and controls. Also homozygotes and heterozygotes for the MICA 5.1 variant gene were not at an increased risk for skin cancer compared to individuals without this variant gene and infection with human papillomavirus did not materially influence these findings. The same group of cases and controls was large enough to show an association between melanocortin 1 receptor gene polymorphism and skin cancer and to reasonably exclude an association between p53 codon 72 polymorphism and skin cancer. Therefore, we conclude that an association between MICA gene polymorphism and nonmelanoma skin cancer is not likely.