202 Health Physics exposure to 157 members of the Utrik community. Lesser exposures of different degrees affected approxi- mately 13,000 other Marshallese then living elsewhere in the archipelago, including the major population centers of Majuro and Kwajalein. The purpose of the present paper is to update and replace the cancer risk estimates of an earlier report (DCEG 2004)titled “Estimation of the Baseline Number of Cancers among Marshallese and the Number of Cancers Attributable to Exposure to Fallout from Nuclear Weapons Testing Conducted in the Marshall Islands” and to provide thorough documentation of the methods used to estimate cancer risks. Presented here, for the first time, is a comprehensive assessment of cancer risks from exposure to fallout from all nuclear tests for all Marshallese alive during the years 1948 through 1970. MATERIALS AND METHODS Ionizing radiation exposure is a known cancerrisk factor and, because it is often possible to estimate tissue-specific radiation doses with reasonable precision, the relationship between dose and subsequent cancerrisk is probably better quantified than for any other common environmental carcinogen. In the companion papers, estimates of organ-specific radiation absorbed doses from fallout-related internal (ingested) and external ra- dioactive materials are derived for residents of different atolls affected by different test explosions, by year and age. These values have been summarized to provide yearly radiation doses correspondingto all calendar years from 1948 through 1970,all historically inhabited atolls, and all ages when exposures occurred. Although small-scale medical studies have been reported describing early and late health effects among residents of Rongelap and Utrik in particular (Conard et al. 1970; Hamilton et al. 1987; Cronkite et al. 1997; Takahashi et al. 1997, 2001), and records are available concerning compensation claims awarded to RMIresidents who developed cancers and other health problems subsequent to fallout-related events (NCT 2004), the infrastructure of medical reporting and records in the RMIis not sufficient to support detailed epidemiological studies such as those carried out among survivors of the atomic bombings of Hiroshima and Nagasaki, Japan (Preston et al. 2003, 2007). Fortunately, however, much of the dose-response information provided by studies of the atomic bomb survivors and of other populations exposed to medical, occupational, and other sources of radiation, is summarized in the recent report of the U.S. National Research Council’s Committee to Assess August 2010, Volume 99, Number 2 Health Risks from Exposure to Low Levels of Ionizing Radiation, otherwise known as BEIR VII (NRC 2006). The goal of the present investigation is to estimate the likely consequences of the nuclear tests in terms of cancer risk to the MI population from fallout-related radiation exposures, and to that end we have used the BEIR VII dose-response models, with a few modifica- tions which are noted in the text. Cancer risk projections are provided for post-1948 lifetime baseline? and radiation-related numbers of leukemias and cancers of the thyroid, stomach, colon, and all remaining solid cancer sites considered as a group. Lifetime risk is further divided into “past” (from 1948 through 2008) and “future” (after 2008) periods. Population Table 1 showstotal MI population numbers bysex, as determined by censuses carried out in 1935, 1958, 1967, and 1973 (RMI 1987, Tables 1.1 and 1.7). These data indicate that the MI population increased by about 33% over the 1935 value between 1935 and 1958, by 36% of the 1958 value between 1958 and 1967, and by another 28% between 1967 and 1973, and that the male/female ratio decreased from 1.10 in 1935 to 1.06 in 1958 and to a little over 1.04 in 1967 and 1973. Fig. 1 shows estimated sex-specific MI population sizes by year, obtained by interpolation using the standard fitted Bezier cubic spline curve algorithm (Foley et al. 1992) as implemented in the Microsoft Excel’ spreadsheet command for XY scatter plot with data points connected by smoothed lines. The census reports of 1958, 1967, and 1973 also gavetotal population numbers byatoll (RMI 1987, Table 1.2), which weused to apportion the total populations for these and other years by atoll. As a general rule, for yearly apportioning of the interpolated yearly total numbers amongthe various atolls, we used the 1958 proportional allocation for each of the years 1948-1957, and interpolated linearly between 1958 and 1967 for 19591966 and between 1967 and 1973 for 1968-1970. Estimated numbers are plotted by year in Fig. 2 for the combined populations of the northern atolls of Rongelap and Utrik, the mid-latitude population center Kwajalein, a group of six other mid-latitude atolls (Ailuk, Likiep, Mejit, Ujelang, Wotho, and Wotje), the southern population center Majuro, and the remaining 13 southern atolls (Ailinglaplap, Arno, Aur, Ebon, Jaluit, Kili, Lae, Lib, Maloelap, Mili, Namorik, Namu, and Ujae). Fig. 2 * Cancers that presumably would have occurred in the absence of exposure to radioactive fallout. * Provided for information purposes only. Identification of software does not imply any endorsement.