The chemistry at the interface between homogeneous catalysis and supramolecular chemistry, namely supramolecular catalysis, has attracted sustained attention over the last decade.[1] In a departure from early endeavors to create preorganized molecular receptors as an artificial enzyme, the current approach toward understanding and mimicking enzymatic catalysis has focused on the design of systems capable of spontaneously generating well-defined, functional supramolecular architectures by self-assembly. The inner spaces of these assemblies serve as an isolated reaction vessel that can confine the substrates, thus delivering otherwise unattainable reactivity and selectivity.[2] In parallel, new possibilities of supramolecular catalysis have been demonstrated by the marriage of multicomponent assemblies with homogeneous transition-metal catalysis and organic molecular catalysis since the seminal studies of Breit, von Leeuwen, and Reek in 2003. [3] We recently reported that chiral P-spiro triaminoiminophosphorane 1 and three equivalents of arylhydroxides (ArOH) spontaneously assembled into the highly organized molecular structure 1·[ArOH] 3 by the formation of an ion-pair-assisted hydrogen-bonding network (Scheme 1). [4,5] Furthermore, the resultant network 1· [ArOH] 3 , and particularly that with a 3,5-dichlorophenol component (3,5-Cl 2 C 6 H 3 OH; 2), exerted efficient cooperative catalysis in the stereoselective conjugate addition of an acyl anion equivalent. Although the discrete three-dimensional structure of 1·[PhOH] 3 in solid state was determined by single-crystal X-ray diffraction analysis, the actual behavior of this type of supramolecular catalyst in solution remains an important issue to be resolved. This situation prompted us to conduct a series of spectroscopic analyses of a solution of 1 and 2 in an organic solvent, which revealed not only the effectiveness of the low-temperature 31 P NMR spectroscopy measurement for tracing the solution structure, but also unexpected, yet intriguing, phenomenon regarding the mode of molecular association. Herein, we disclose the stepwise and exclusive generation of three types of molecular assemblies, 1·[2] n (n = 1-3), in solution by simply adjusting the stoichiometry of 2. Each structure was unequivocally verified in the solid state by X-ray crystallographic analysis. The finding that the mode of the spontaneous assembly of 1 and 2 can be precisely controlled in solution suggests the possibility of selective use of 1·[ArOH] n (n = 1-3) as a requisite catalyst for target organic transformations and could also provide a expedient means to gain insights into the structural integrity of the reactive intermediate.Initially, 31