Abstract The soybean system used for detecting environmental mutagens is analyzed for various types of spots on the leaves of heterozygous Y 11y 11 plants and homozygous Y 11y 11's induced by a nitrosoamine (dimethyl nitrosoamine, DMN) and a nitrosoamide (methyl nitrosourea, MNU). It is shown that the nitrosoamine can be “activated” by the seed (is converted to a true mutagen) without the addition of NADPH or S-9 fraction of the liver homogenate as is necessary in animal tissue culture or bacterial studies. Whereas somatic mosaicism in soybean can be induced with a dose as low as 1.25 ppm of DMN, the upper limit in spot production is reached at around 60 ppm concentration, applied for 0–24 h. Such saturation effect may be due to a limited amount of DMN being converted to true mutagen. MNU, on the other hand, does not show such limitations, perhaps because of its property of being a direct mutagen not necessitating an intermediate step required for converting the promutagen DMN. The frequency of twin spots on Y 11y 11 leaves increases only slightly by either DMN or MNU, suggesting only a small increase in somatic crossing-over induced by the two chemicals. The yellow spots increase the most, perhaps due to segmental losses carrying Y 11 or non-complementary segregation of exchanges involving non-homologous chromosomes. Neither chemical is found capable of mutating y 11 to Y 11 as seen by the general lack of light green sectors on y 11Y 11 plants. Usefulness of the soybean system in studying mutagenesis is briefly discussed.