The development of highly active enzyme-based biocatalysts for synthetic applications in organic solvents represents a challenge for biomolecular researchers and engineers. Enzymatic catalysis in non-aqueous media can be useful when substrates or products have limited or poor solubility in water or if water is a reaction product. Improved thermal stability of enzymes in non-aqueous media and some advantageous effects on enantioselectivity have also been reported. It is however imperative to find inexpensive and simple immobilization techniques since enzymes are generally not soluble in solvents that do not denature them. The immobilized enzyme preparations needed for use in solvents should be easy to prepare, based on widely available materials, and should make excellent use of the catalytic competency of the enzyme. Our strategy comprises the immobilization of enzymes on fumed silica nanoparticles by lyophilization. The catalytic activity of Candida antarctica Lipase B (CALB) in hexane for a model esterification is comparable with that of the commercial preparation Novozym 435®. The immobilization is basically achieved in two sequential stages: (i) physical adsorption of the enzyme from aqueous suspension on fumed silica nanoparticles, and (ii) lyophilization. The kinetics of enzyme adsorption as a function of the fumed silica charge and pH of the aqueous phase before lyophilization will be discussed. The catalytic activity was assessed from initial reaction rate experiments and the physical enzyme distribution on the support was investigated by confocal microscopy imaging. Additionally, we explored the thermal and storage stability of our preparations. In light of our previous work with subtilisin Carlsberg our results with CALB are somewhat surprising since a counterintuitive loss of activity at very high fumed silica concentrations was detected. We hypothesize based on our data and recent models put forth by the research group of Professor Gross (Polytechnic University) that this apparent optimum enzyme-to-support ratio may be due to the unique protein-protein and surface-protein interactions during the adsorption of CALB on fumed silica.