Enzymes/Nanoparticles (NPs) bioconjugates are massively used nowadays to develop thin films for optical and electrochemical biosensors. Nevertheless, their full characterization as a thin coating onto electrodes remains little discussed, in particular the influence of NPs size and enzyme/NPs ratio used in the electrodeposition solution. In this study, GOx (160 kDa) and HRP (44 kDa) were used in association with tannic acid capped gold NPs (a series with sizes from 7 to 40 nm) to electrodeposit biosensor coatings, sensitive towards glucose and H2O2, respectively. The electrodeposition process was based on a mussel-inspired electro-crosslinking between gallol moieties of tannic acid (at the surface of NPs) and amine moieties of the enzymes. On one hand, the sensitivity of the GOx/NPs coatings depends strongly on the NP size and the enzyme/NPs molar ratio of the electrodeposition solution. An optimal sensitivity was obtained by electrodeposition of 11 nm NPs at a GOx/NPs molar ratio close to the theoretical value of the enzyme monolayer. On the other hand, a modest influence of the NPs size was found on the sensitivity in the case of the electrodeposited HRP/NPs coatings, reaching a plateau at the HRP/NPs molar ratio close to the value of the theoretical enzyme monolayer. In both cases, the enzyme/NPs molar ratio played a role in the sensitivity. To fully understand the parameters driving the biosensor sensitivity, a comprehensive evaluation of the colloidal state of the bioconjugates is proposed here.