Soy isoflavones have anticarcinogenic, antioxidant, and antiatherosclerotic activities. They also interact with the estrogen receptor, which makes them weak or moderate phytoestrogens. Because of their bioactivity, isoflavone bioavailability has been extensively studied in humans. This review summarizes data from intervention studies in humans, focusing on the factors that affect bioavailability. Summarizing data from 16 studies shows that the maximum concentration in plasma normalized to a constant dose of genistin is approximately 1.6 times that of genistein, and daidzin is approximately 1.8-fold higher than daidzein, whereas the half-life is not significantly different for aglycone and glucoside. There is a wide variation in the reported percentage urinary excretion that is not dependent on dose. Bioavailability is increased by a rapid gut transit time and by low fecal digestion rates and is decreased by a fiber-rich diet. There is no difference in bioavailability between pre- and postmenopausal women. The daily ingestion of soymilk for 1 wk does not affect bioavailability, but daily ingestion for a month increases excretion of equol in women. The factors or habitual diet characteristics that influence equol production are not clear, but equol production is limited with an immature flora. There is no consensus on which source of isoflavones results in the highest isoflavone bioavailability, and published studies present different results, although bioavailability is affected by whether the dose is given as food or drink. In conclusion, it is important to consider the factors affecting bioavailability of isoflavones when designing intervention studies.