The thesis is based on 13 publications in English and a review of the literature. The underlying work was done with the overall aim to describe incidence patterns for anal squamous carcinoma (anal SC) and to contribute new insight into the causes of this neoplasm. The work, supported by the Danish Cancer Society, was carried out in the period 1991-2000 while I was employed at 1) the Danish Cancer Registry, 2) Statens Serum Institute, Department of Epidemiology Research, and 3) the National Cancer Institute, Viral Epidemiology Branch, Maryland. Study designs employed include a) population-based incidence studies in Denmark and the United States, b) register-based case-control studies and cohort studies for the scrutiny of multiple cancer patterns among patients with anal SC and for the study of anal SC risk among individuals with certain non-malignant diseases of the anal region as well as among persons with the acquired immuno-deficiency syndrome (AIDS), c) a nationwide interview-based case-control study of risk factors for anal SC and in Denmark and Sweden, and d) a combined molecular biological and histological analysis examining the association of human papillomavirus (HPV) status with histopathological and anatomical characteristics in anal SC tissues. The epidemiology of anal SC has changed remarkably during the second half of the 20th century. In Denmark, age-adjusted incidence rates per 100,000 person-years increased during the period 1943-1997 from around 0.2 among both men an women to 0.5 among men and 1.0 among women. Where systematically studied, incidence rates of anal SC have also been found to increase in a few other countries (Sweden and the United States). Register-based multiple cancer studies have shown an excess of previous and subsequent genital cancers of squamous histology among women with anal SC. This is likely to reflect common susceptibility toward infection with cancer-associated HPV types shared by all anogenital organs covered by squamous epithelium. A study based on data from the United States supports the possible role of anogenital SC-associated HPV types in the development of some tonsillar cancers. Observations in other register-based investigations and in the Danish-Swedish case-control study challenge the long held belief that benign anal lesions (e.g. hemorrhoids) and anal inflammation (e.g. in association with Crohn's disease) are linked to an increased risk of anal SC. The work documents strong links between a variety of sexual behavior measures and the risk of anal SC. The previously observed excess of anal SC among homosexual men is confirmed. Unlike in previous studies, an increased risk of anal SC associated with promiscuous heterosexual activity is also documented among both men and women. Measures of sexual extroversion, rather than sexual preference, are linked to the risk of anal SC presumably by means of higher rates of anal HPV infection in people with such behaviors. A new hypothesis is proposed to explain how smoking, the only consistently observed non-sexual risk factor in previous studies, might be associated with anal SC risk. A study of the short-term cancer profile among 309.365 AIDS patients in the United States provided no evidence to support immunosuppression as a major risk factor for anal SC in the first two years after the AIDS diagnosis. With the introduction around 1996 of highly active anti-retroviral therapy regimens, however, the future may hold a para-doxical increase in the incidence of anal SC along with the increased life expectancy in this population. Examination of tumor tissues from patients in Denmark and Sweden by the polymerase chain reaction techique showed a high proportion of anal SCs to be positive for types of HPV that are associated with high risk of cervical cancer (90%, 100%, and 58% among women, homosexual men, and heterosexual men, respectively). Tumor tissues from control subjects with adenocarcinoma of the rectum were consistently HPV-negative. A combined molecular biological and histological analysis showed that anal SCs with likely origin in the anal canal are 7.5 times more likely to be HPV-positive than anal SCs with likely origin in the perianal skin. Additionally, cancer of the anal canal are more likely than those of the perianal skin to be characterized histologically by small or medium-sized tumor cells, basaloid features, and little or no keratinization. Epidemiological studies from the past two decades have contributed imprtantly to our current understanding of anal SC and its causes. Most cases of this neoplasm can now be considered as a consequence of sexually or otherwise acquired infection in the anal mucosa with types of HPV already known to be involved in cancers of the uterine cervix. Expectedly, the upward trend in the incidence of anal SC seen over the past half century will continue in many years to come. However, there is currently widespread, yet cautious, optimism regarding the prospects for a prophylactic HPV vaccine. It this optimism is justified, the future may bring drastic reductions in the incidence of HPV-associated morbidities, including SC.