The utility of computational study lies not only in rationalizing a chemical phenomenon but also in its predictive value. Broadly, the scope of my research work includes understanding of the structure and bonding of molecules as well as reaction mechanisms using computational techniques. Here I will discuss three research problems where computational results successfully rationalize and predict the experimental outcome. Firstly, we will describe the electronic structure and bonding of all the possible cyclic isomers of B2AlHnm (where n =3D 3 =96 6 and m =3D -2 to 1) = which is isoelectronic to the cyclopropenyl cation.1 A comparative study among all the isomers of homocyclic and heterocyclic three- membered boron and aluminum hydrides has also been done to understand the factors that differentiate their hydride chemistry. We will also discuss about two different approaches to stabilize neutral planar B3R3 rings. In a mechanistic study, we have designed a a priori system which can undergo two competing biradical generating processes, namely the Myers-Saito (MS) and Garratt-Braverman (GB) Cyclizations.2,3 We will present a detailed mechanistic study of both the reactions, which indicates the preference of the GB cyclization over MS cyclization. The theoretical prediction is in agreement with the experimental findings. We will also describe a conformational constraint-based strategy to switch the selectivity from GB to MS/Schmittel pathway.4 In another study, we will talk about a DFT study to illustrate the effect of the a) solvent, b) ancillary ligand, (L) c) leaving group, (Hal) and d) metal (M) on the equilibrium between metal boryl (1) and borylene (2) complexes (Scheme 1).5,6