OBJECTIVE The primary purpose of this study was to examine the occurrence of cancer in Alaska Native (AN) children (under age 20). Although several studies have compared differences in cancer incidence between white and black children, few have examined cancer among Alaska Natives/American Indians. We know of no published article describing cancer incidence in AN children. We compared our findings with those of American Indian children of New Mexico and of Alaska white children. Data on mortality, survival, and prevalence are also included. Alaska Native is the term used collectively for the inhabitants whose ancestors occupied the area before European contact of what is now the state of Alaska. Alaska Natives include Eskimo, Indian, and Aleut groups. Although the 3 major groups differ in culture, language, and probably genetics, there are similarities in numerous social and economic indicators. The Northern Eskimo of Alaska (Inupiat) are related to Canadian and Greenland Inuit. Indians in Alaska include Athabaskan (in the interior of the state), who share commonalities with Canadian Athabaskan as well as with Navajo and Apache in the southwestern United States. Tlingit, Haida, and Tsimshian groups reside primarily in the southeast panhandle of the state. The panhandle Indian groups are similar to those of British Columbia. METHODS Data on cancer incidence are from the Alaska Native Tumor Registry, 1969-1996. We studied children under age 20 to make our results comparable to national data as presented in the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) Pediatric Monograph. Population data for AN are based on census data and Indian Health Service intercensal estimates. Data for US whites and New Mexico Indians are from the National Cancer Institute's SEER program. Calculations were made using SEERStat software. Data for Alaska whites are for the years 1996-2000. (The Alaska Cancer Registry has collected data for all Alaskans only since 1996). Odds ratios (ORs) of rates with 95% confidence intervals (CIs) were calculated. RESULTS The rate among all AN children (both sexes) for all cancers combined is similar to that of US whites (OR: 1.0; 95% CI: 0.8-1.1). Examination of childhood cancer rates by ethnicity, however, reveal that rates are significantly lower for Indian (OR: 0.6; 95% CI: 0.4-0.8) but not significantly different for Eskimo or Aleut children. For most International Classification of Childhood Cancers groups, incidence rates for AN children are also similar to those of US whites. However, AN children are at significantly higher risk for hepatic tumors (OR: 13.1; 95% CI: 7.9-20.5), particularly hepatocellular carcinoma (OR: 43.8; 95% CI:24.4-75.1) and retinoblastoma (OR: 2.8; 95% CI: 1.3-5.3). By ethnic group, rates for hepatocellular carcinoma are significantly high only for Eskimo. Rates for all AN children are lower for neuroblastoma (OR: 0.1; 95% CI: 0.1-0.6) and lymphoma (OR: 0.5; 95% CI: 0.3-0.9), particularly Hodgkin's disease (OR: 0.2; 95% CI: 0.0-0.5). On the basis of 5 years of data, rates for Alaska white children do not seem to differ from those of US white children. Because of our findings of differences between AN and US whites, we reviewed data of other relevant populations, specifically American Indian data from the New Mexico SEER registry. Using SEER data and SEER software, we calculated rates for New Mexican American Indians (NMAI) and compared them with US white rates. Rates for all cancers combined among NMAI are significantly lower than for US white (OR: 0.8). However, similar to AN children, the rate among NMAI for retinoblastoma is higher compared with US whites (OR: 2.5; 95% CI:1.4-4.5). Similar to AN, NMAI also seem to be at low risk for neuroblastoma (OR: 0.2; 95% CI: 0.1-0.7), lymphoma as a group (OR: 0.1; 95% CI: 0.0-0.3), and, specifically, Hodgkin's disease (OR: 0.1; 95% CI: 0.0-0.4). Rates among NMAI children are low for central nervous system tumors (OR: 0.5; 95% CI: 0.3-0.7). The average annual age-adjusted cancer mortality rate among AN children is lower but not significantly lower than that of US white children (28.6 vs 37.3 per million). CONCLUSIONS Comparison of AN rates for all cancers combined are similar to those of US and Alaska white children but seem higher than those of NMAI. Differences between AN and US whites exist for select International Classification of Childhood Cancers groups. The most striking rate differences are found in hepatic tumors, largely because of elevated rates of hepatitis B-associated hepatocellular carcinoma. All children in our study with hepatocellular carcinoma were hepatitis B antigen positive. A statewide hepatitis B virus immunization program was begun in late 1982. Although 16 children who were born before 1983 developed hepatocellular carcinoma, no children who were born in the 20 years since hepatitis B immunization was instituted among infants have received a diagnosis of hepatocellular carcinoma, a significant difference. Comparing AN and US white childhood cancer rates after removing hepatocellular carcinoma cases from both populations results in an OR of 0.8 (95% CI: 0.7-1.0). Thus, if no increase in other childhood cancers occurs in the coming generations, then rates for childhood cancer may soon be significantly lower than those in US white children. Rates are low for all lymphomas, largely because of very low rates of Hodgkin's disease. Rates are also low for neuroblastoma. It is reassuring that rates for AN children are not in excess and do not seem to be increasing. There is concern among the population regarding environmental exposure, including ionizing radiation. Our data do not show excess childhood leukemia or thyroid cancers, malignancies for which radiation is known to increase risk.