Liquid crystals form a subclass of soft materials which is easily influenced and deformed by a surface, an interface and the geometry. Of particular interest was the confinement of liquid crystals in shell geometry, imposing real or virtual defects that the liquid crystal cannot avoid. With the help of microfluidics, we prepared our research platform, liquid crystal shells, which contain and are surrounded by aqueous phases. In order to maintain such a shell structure in the aqueous phases, immiscible with the liquid crystal, appropriate stabilization is required. Here I explored two different pathways of interfacial stabilization and polymer stabilization and their impact on liquid crystal self-assembly.We primarily used either a polymeric or an ionic surfactant dissolving in water to stabilize shells and tune boundary conditions of shells. Depending on symmetrically or asymmetrically imposed boundary conditions, the nematic–isotropic phase transition appears as a single transition or separated into two steps. The latter phenomenon can be understood as a result of an ordering-enhancing effect by surfactants. The nematic–smectic A phase transition was also investigated under varying boundary conditions. With a precise temperature control, we explored equilibrium smectic structures and introduce a new arrangement of focal conic arrays in shell geometry. Beyond stabilizing the shell from the shell exterior, we also incorporated a photosensitive surface agent within the shell, enabling dynamic and reversible photoswitching of the liquid crystal alignment in real time.However, shells with interfacial stabilization cannot survive more than several weeks due to their intrinsic fluid interfaces. In particular, a liquid crystal shell can serve as a permeable membrane which lets the constituents of aqueous phases pass through, giving a significant influence on the liquid crystalline order. To tame liquid crystal self-assembly and make the shell structure permanent, we used photopolymerization to stabilize the shells. With only 5% monomer, the entire configuration of each liquid crystal shell was locked and shell lifetime extended beyond several months. The liquid crystalline order was visualized on the nanoscale via the polymer network and we further demonstrated that the shell configurations can be a unique template for creating complex polymer networks. Finally a new experimental approach was introduced to making ultrathin shells and several issues on shell instability and alignment determination were addressed.