The scalability of on-chip analytical systems based on field-effect transistors as pH sensors is limited by the degradation of the signal to noise ratio when decreasing the device size. Nano-sized tri-gate transistors such as silicon nanowires and nanoribbons exhibit improved coupling with the electrolyte environment thanks to their tri dimensional structure. Even though in the framework of solid-state tri-dimensional transistors an enhancement of the channel con-ductivity is achieved by reducing the device width, so far the sensitivity performance of liquid-gate devices has been mostly considered in terms of threshold voltage readout. In pH sensing applications, however, the threshold voltage change is independent on the device size, thus overlooking the potential advantage derived from the improved electrical properties of nano sized devices. In this thesis, sensitivity characterization coupled with a comprehensive noise analysis has been performed on devices ranging from 50 nm to 70 µm in width and from 350 nm to 4250 nm in length. Such comprehensive characterization is made possible thanks to the employement indus-trial top down CMOS compatible technology, which ensures high control over process variation and enables the fabrication of nanowires with a wide range of widths and lengths. The results show that reducing the width below few hundreds of nanometers results in an increase of the conductivity properties of silicon nanoribbons, which ultimately improves the sensitivity with respect to surface charge, hence to pH. This enhanced sensitivity of nanoscaled devices can be further exploited within a multi wire configuration, in which the exposed surface is increased and the noise reduced. We provide experimental evidence that multi wire devices maximize the signal to noise ratio achieving, in the presented technology, a resolution of 0.0028 pH·µm2. This result could have a great impact on the improvement of the scalability of on-chip analytical systems requiring high pH resolution, such as DNA sequencing and quantitative PCR.