An exciting challenge in contemporary science is learning how to make ordered molecular materials with predetermined structures and properties. The effort to reach this goal continues to yield new knowledge of both fundamental and practical value. Of particular interest are ordered molecular materials produced spontaneously by methods of self-assembly, such as crystallization. Although detailed predictions of the structure of molecular crystals remain unreliable, [1] crystal engineering provides increasingly effective strategies for identifying compounds predisposed to crystallize in particular ways. [2] One such strategy, which has been called molecular tectonics, [3,4] relies on special compounds that participate in multiple directional interactions. Such compounds, called tectons from the Greek word for builder, are programmed to associate and to form networks in which each molecule is placed in a predictable position with respect to its neighbors. Herein, we show how two different directional forces, hydrogen bonding and dipole-dipole interactions, [5] can be used in potent combination to engineer molecular networks with predictable cleavage and high porosity. [6] Tetrahedral tectons 1 with peripheral sites of association (*) are predisposed to form three-dimensional four-connected networks, whereas trigonal analogues 2 favor sheets (Figure 1). In the first case, the entire three-dimensional structure is programmed by the tectons and maintained by strong interactions in all directions; in the second case, however, the ultimate structure must result from the stacking or interpenetration of sheets, [7] governed by forces that are not normally subject to rational control. Examples of structural polytypism resulting from subtle variations in the stacking of sheets are numerous in chemistry and include the diverse phases of graphite, [8] sheet silicates in which the layers have different offsets, [9] and various hydrogen-bonded molecular sheets. [10] For many applications, polytypism is undesirable because it makes full structural predictions unreliable, even when the positions of all the atoms in individual sheets are well established.