The purpose of this study was to design and develop a dynamic culture system for stem cells inside a bubble-confined cell array. Our goal was to recreate an environment better mimicking in vivo conditions of natural tissues while meeting the requirements of important processes such as pharmacological screening tests. Cell cultures were supported by polyacrylamide based hydrogels; such hydrogels proved to be biocompatible, optically transparent, extremely hydrophilic while being non adhesive toward cells. Hydrogels shape was conditioned using selective photopolymerization techniques. Three independent gel circles were created on a pre-treated 75x25 microscope slides; glass surfaces were functionalized by covalent binding with a hydrophobic thin layer of methacrylate groups. The cell array was created deposing micrometer-scale protein islands on the hydrogels surfaces, thus allowing generation of selective adhesion sites. Proteins were dropped using a robotic DNA microarrayer and following a well defined pattern. We tested different proteins (laminin, collagen I, polylisine) in concentrations ranging from 100 to 500 microL/mL. The deposition of FITC-conjugated proteins was used to accurately control the spots position and dimensions and to roughly evaluate the protein adsorption profiles and concentration at the hydrogel surface through fluorescence image analysis. Studies were performed in order to assess the growth and proliferation of stem cells seeded and cultured on such protein patterned hydrogel surfaces both under static and dynamic culture conditions. All procedures were optimized and resulted in a highly selective adhesion, spreading and proliferation of cells within spotted islands while hydrogel non-coated surfaces prevented cells adhesion. The dynamic culture apparatus was designed in order to assess a stable and constant medium flow through our array, ensuring a continuous supply of nutrients and removal of wastes for our cell cultures. We evaluated the shear stress levels in the dynamic system running fluodynamic simulations modeled on our bubble and used operative parameters. Such values proved to be much lower than those creating negative effects on cell cultures.