The invention relates to a method for identifying an estrus in a female, non-human mammal within a group of animals of the same kind, wherein, in a first method step, a parameter specific for the movement activity of the animal in dependence on the time is detected and the animal is selected for further evaluation, if said parameter lies above the average value of the parameter specific for the movement activity in the corresponding animal for a predetermined length of time. Similar methods which are currently known are unreliable and dependent on random events. The problem of designing a corresponding method which ensures a high degree of reliability is solved according to the invention in that, in a second method step, a parameter specific for the group formation of the animal by at least one additional animal of the same kind from the group is determined for the same length of time and, in a third method step, an output for the probability of an estrus of the animal in dependence on the value of the parameter specific for pair formation is produced.
