The study of multiple female mating preferences and multiple male signals r
equires correct and precise measurement of preferences. A review is given o
f existing preference test paradigms. Non-interactive preference tests usin
g computer animations perfectly fulfil the demands for the study of multipl
e preferences for visual traits: exclusion of confounding variables, exclus
ion of variation within and between male pairs, great potential of experime
ntal manipulation of single and combinations of visual traits including beh
aviour. We give a detailed description for the production oi. computer anim
ation movies based an commercial software. Finally, we show how computer an
imations can be properly applied to the testing of mating preferences. In s
ticklebacks, female mating preferences that were tested in this way agreed
with preferences that were measured with other test paradigms.