The inherent properties of momentary time sampling (MTS) and partial interval recording (PIR) are examined. Findings derived from computer simulation investigations are discussed in terms of the mode of operation of the two time-sampling techniques. It is seen that the advantage of MTS is that it can, under certain restricted circumstances, estimate absolute duration of behaviour occurring. The important disadvantage of MTS is that it is relatively insensitive when estimating degree of change of behaviour. In contrast, although PIR cannot accurately measure absolute duration it is more sensitive to behaviour change than is MTS. It is concluded that the practitioner who wishes to use one of these methods of time sampling must carefully consider the aims and possible effects of the investigation before deciding which method to use.