Understanding electron-transfer (ET) processes in dye-sensitized solar cells (DSSCs) is crucial to improving their device performance. Recently, covalent attachment of dye molecules to mesoporous semiconductor nanoparticle films via molecular linkers has been employed to increase the stability of DSSC photoanodes. The power conversion efficiency (PCE) of these DSSCs, however, is lower than DSSCs with conventional unmodified photoanodes in this study. Ultrafast transient absorption pump-probe spectroscopy (TAPPS) has been used to study the electron injection process from N719 dye molecules to TiO2 nanoparticles (NPs) in DSSC photoanodes with and without the presence of two silane-based linker molecules: 3-aminopropyltriethoxysilane (APTES) and p-aminophenyltrimethoxysilane (APhS). Ultrafast biphasic electron injection kinetics were observed in all three photoanodes using a 530 nm pump wavelength and 860 nm probe wavelength. Both the slow and fast decay components, attributed to electron injection from singlet and triplet excited states, respectively, of the N719 dye to the TiO2 conduction band, are hindered by the molecular linkers. The hindering effect is less significant with the APhS linker than the APTES linker and is more significant for the singlet-state channel than the triplet-state one. Electron injection from the vibrationally excited states is less affected by the linkers. The spectroscopic results are interpreted on the basis of the standard ET theory and can be used to guide selection of molecular linkers for DSSCs with better device performance. Other factors that affect the efficiency and stability of the DSSCs are also discussed. The relatively lower PCE of the covalently attached photoanodes is attributed to the multilayer and aggregation of the dye molecules as well as the linkers.