The sol–gel method is employed for producing high surface area silica nanoparticles from a cheap precursor, that is water glass (sodium silicate). To design the experiments systematically, the response surface method (RSM) combined with the central composite design (CCD) approach is used. Four major factors including the concentration of sodium silicate solution, solution pH, reaction temperature and reaction time are identified as the major controlling parameters and the particle surface area is considered as the response or the output parameter. A total of 31 experiments are designated by the CCD. The experiments are conducted at a centre point chosen based on experience and at its vicinity to investigate how the response changes as the factors change. Nanoparticles with a surface area as high as 630 m2/g and a particle size as low as 8 nm are produced at the optimum parameters of sodium silicate concentration of 1.5 × 10−4 g/L, pH of 4, a reaction temperature of 25°C and a reaction time of 1.5 h. A regression analysis is performed on the experimental data and a correlation is obtained that may be used to predict the particle surface area and investigate the effect of varying the factors on the response. Within the range of parameters studied here, it is found that the solution pH and then the process temperature have a profound effect on the surface area and concentration and reaction time have a moderate effect. An increase of the solution pH from 4 to higher values results in a rapid drop in the particle surface area.