AbstractThe utility, vehicle required, the time needed for assembly and relocation and the costs associated with six different rainfall simulators were studied. The results can be summarized as follows:1The simulator designed byKainz&Eicher(1990) is the least expensive to operate. This simulator also has excellent experimental versatility. Its only restriction is a short slope length.2The rainfall simulator ofKarl&Toldrian(1973) is also inexpensive to operate, but because of disadvantages such as low falling height, wind susceptibility, and the need for more workers than other simulators, it should be used mainly where other simulators are not practical (forests, steep slopes, or long slopes).3The rainfall simulator ofRichter et al.(1988) has limited mobility in the field and is highly susceptible to wind drift. Moreover, its narrow plots do not cover a representative unit of a cultivated field.4The modifiedSwanson‐type simulator(Auerswald, (1986) is suitable for large plots if a sufficient water supply is available.5The simulator ofKromer&Vöhringer(1988) has the highest fixed costs, but labour is used efficiently.6The simulator ofSchmidt(1983) can only be considered for stationary use because of its long assembly time. Other disadvantages are narrow plot size and high wind susceptibility.Depending on the selection of the simulator, total costs of between 400 DM and 2000 DM per plot should be expected even with high use rat