Multigeneration tests are very useful for the assessment of long term toxicity of pollutants such as endocrine disruptor compounds. In this study, multigeneration reproduction tests adapted from the ISO standard 11267 were conducted with the Collembola Folsomia candida. Springtails were exposed to artificial soil contaminated with four insect growth regulators (methoprene, fenoxycarb, teflubenzuron, and precocene II) according to two different experimental set-ups. In the first set-up, the parental generation (F(0)) of Collembola was exposed to a pollutant for 28 days. Juveniles from the F(1) generation were transferred to uncontaminated soil for another 28-day period to generate the F(2) generation. In the second set-up, the F(0) generation was exposed to a pollutant for 10 days before being transferred to uncontaminated soil to reproduce. After 18-28 days, juveniles from the F(1) were transferred to clean soil to generate the F(2) generation. An effect on the number of hatched juveniles of the F(2) generation was observed for methoprene after exposure of the F(0) for 28 days and hatching of F(1) in contaminated soil. For methoprene and teflubenzuron, significant effects were even observed on the F(2) generation with the second experimental set-up, when only the F(0) generation was exposed for 10 days. This shows that the impact of these substances is transgenerational, which can have important consequences for the population of these or other organisms. No effect on the F(2) generation was observed with fenoxycarb and precocene II with the 10-day exposure experiment. Our results show that the developed experimental procedures are appropriate to assess the long term effects of endocrine disrupting compounds on the reproduction of the non-target species F. candida. Another important finding is that two substances with the same predicted mode of action (i.e., the two juvenile hormone analogues fenoxycarb and methoprene) do not necessarily affect the same endpoints in F. candida.