A comparison is made, between three methods for visualizing the Minkowski space, in view of their application to relativistic kinematics.The velocity manifold method, in which a vector is represented by a point on the upper sheet of the unit-radius hyperboloid, presents two advantages: there is no distortion of angles, and the method is independent of the choice of any particular frame or observer. It is, however, limited to the representation of positive timelike vectors.The Cayley map method allows the representation of spacelike as well as timelike vectors, but it has two drawbacks: it distorts the angles, and it depends on the choice of a reference system.In the third method, the use of a unified space–time formalism permits one to visualize any kind of vector. This method generalizes the velocity manifold method, and presents the same two advantages: it provides us with a conformal representation, which is independent of any particular observer.We also present a rather concise comparison between two covariant methods of calculation, considered from the point of view of relativistic kinematics. One method is the familiar tensor calculus; the other is the spinor calculus, based on the representation of four-vectors by matrices.