We evaluated the methods and results of nine cohort studies dealing with the biological effects on human health from exposure to radiofrequencies/microwaves, published between 1980 and 2002. The size of the cohorts varied between 304 (3,362 person years) and nearly 200,000 persons (2.7 million person years). As exposures were defined: dielectric heaters in a plastic manufacturing plant, working with radio devices (professional and amateur), production of wireless communication technologies, radar devices of the Canadian police, radar units used by the military as well as artificially produced electromagnetic pulses similar to those after a nuclear explosion. In all studies (except one that used a qualitative job-exposure-matrix) either the duration of occupational work as an approximation to actual exposure was determined or a simple yes/no differentiation was used based on a definition of high-exposed and/or low-exposed (occupational) groups. Either total mortality, cancer mortality, cancer incidence or other outcomes were estimated. In most of the studies, an increased risk for various types of cancer was found in exposed study participants, although in different organs. The overall results were, however, inconsistent. The most important limitations of the studies were the lack of measurements referring to past and current exposures and, thus, the unknown details on actual exposure, the use of possibly biased data as well as the lack of adjustment for potential confounders and the use of indirect standardization techniques. Due to these limitations and the inconsistencies of the results it has to be concluded that the studies give no evidence of high frequency emissions causing cancer.