Management. Can. 1. Fish. Aqrmat. Sci. 44(Suppl. 2): 75-83.Optimal control theory is a useful tool for analysis of general mathematical models of fisheries and harvest strategies. It provides a framework for management decisions, but is too general for direct application to many specific management problems. A review of applications of optimal control theory to fisheries management is presented, with a general audience in mind. Advantages and shortcomings of optimal control theory in the context of fisheries management are discussed with some specific examples.La theorie du contr6le optimal constitue un outil utile pour I'analysede modeles mathematiques g6neraux des peches et des strategies d'exploitation. Elle fournit une base pour les dkcisions relatives a la gestion mais elle est trop generale pour 6tre appliquee directement a d e nornbreux probl&mes d e gestion determines. L'armteur presente, a I'intention du public, un examen des applications de cette theorie a la gestion haliermtique. II illustre a I'aide d'exemples ses avantages et ses inconvenients dans le contexte de la gestion haiieutique.In vector notation, (1) becomes wherex(t) is an n x 1 vector, representing the state of the system (e.g. species density), u(t) is an r x 1 input or control vector (e.g. harvest quota or harvest rate), and f is a vector-valued function. With time delays T, we have and for discrete systems, where k is the kth time interval.The optimal control problem is stated as follows: Given the state eq. (I), a set of boundary conditions on the state variables at the initial and terminal times, and a set of constraints on the state and control variables, determine the admissible controls u(t) so that a performance index (cost function) is minimized (or maximized). The boundiary conditions are given by x(to) and SEX($)] where to, 9, and S denote, respectively, the initial time, the final time, and the target set. Note that to is always fixed, but may be part sf the optimal control problem; e.g. one may strive to minimize rf:The performance index is expressed a% a scalar quantity where 6 and F are scalar-valued functions.The main theoretical approaches to the optimal-control-Can. J . Fish. Aquar. Sci., Vo%. 44, I987 7% Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by McMaster University on 11/25/14 For personal use only. Can. J . Fbh. A q u t . Sei., V01.44, 1987 Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by McMaster University on 11/25/14 For personal use only. (May 1974). When the ecosystem model is not large (e.g . m < 5), and when the parameters of F in (8) are not known, the Routh-Hunwitz stability criterion may be used to establish Can, 9. Fish. Aquat. Sci., VoI. 44, 1987 Can. J. Fish. Aquat. Sci. Downloaded from www.nrcresearchpress.com by McMaster University on 11/25/14For personal use only.