In many fields of study there occur systems of lagged dependences which can be described approximately by systems of linear stationary stochastic differential equations. In this paper we consider the estimation of parameters in such systems, using an approach similar to that of Bartlett (1946, 1955) and Quenouille (1957), and discuss conditions for identification of the separate behaviour relationships of a system. Little is known about the small-sample properties of estimates in systems of this kind, but sampling experiments are being carried out and the results will be published later.