This paper investigates the stability of the sets of efficient points of vectorvalued and set-valued optimization problems when the control cone is regular but need not have a bounded base in the sense of Painleve-Kuratowski convergence of set sequence.

As an application, an existence result of efficient point for the nuclear cone in Hausdorff topological vector space is investigated, which is a generalization of the result in locally convex space.

It raises an idea for sieving multiple objective-programming to find a group of efficient points and to approach the whole of the efficient points.

Using the particular property of cones, we prove that the set of proper efficientpoints is dense in the set of efficient points.

Instead of the classical Hahn-Banach separation theorem, we applythe separation results of Gerstewitz s functional to scalarizations of efficient points, so weneed not require that the feasible sets are convex sets.

We prove that strictly efficient points are indeed equivalent to Henig proper efficient points.

In this paper, a few optimatity conditions for strictly efficient points of set valued optimization are presented by using the concept of contingent derivatives of set valued mad.

Particularly, we show that in reasonable settings the strictly efficient points of a set are dense in the efficient points.