Although the concept of absorbed dose is commonly used in radiation biology as a parameter for comparing the toxic effect of different levels of radiation on a system, there are situations where the absorbed dose by itself is inadequate, and additional dose distribution information is required to explain the observed biological effects. A good example is the irradiation of cells by alpha-particles. This paper reports the use of internal microdosimetry techniques to reinvestigate the dosimetry to two very similar experiments with apparently contradictory dose-response results. Yields of dicentric chromosome aberrations induced in human blood lymphocytes following in vitro exposure to dissolved americium or plutonium at two separate laboratories produced linear dose-response functions, but the slopes of the best-fit straight lines differed by a factor of 12. Our microdosimetric analysis showed the results of one experiment to be inconsistent with a uniform distribution of activity. It also showed that the difference in slope could be attributed to differences in particulate size and spatial distribution as a result of dissimilarities in procedures used for preparing the actinide solutions.