Electron scattering measurements have been made in order to determine the longitudinal and transverse form factors of low-lying level in 10B. With the exception of the broad 5.18 MeV level, results were obtained for all known levels up to 6.7 MeV. The measurements span the momentum transfer range q=0.48–2.58 fm-1. The primary objective of this work was to improve the data on the pure isovector M3 form factor of the 1.740 MeV excitation, the transform of which yields the 1p3/2 single-nucleon wave function. A Woods-Saxon potential was found to provide a much better representation of the data than the harmonic oscillator model, and the rms size of this orbit was determined to be 2.79±0.11 fm in the relative core-particle coordinate frame. Nevertheless, confidence in the quantitative details of this interpretation is hindered by conflicting evidence regarding the contribution of core polarization. Our analysis of the Coulomb elastic form factor gave an rms radius for the ground-state charge distribution equal to 2.58±0.05±0.05 fm, slightly larger than values previously published. Longitudinal and transverse form factors deduced for inelastic transitions were compared with theoretical results of conventional 1p-shell models, models with 1ħω and 2ħω configurations involving the 1s, 2s1d, and 2p1f shells, and finally, a model that included core polarization. Although restricted 1p-shell models were found to provide good predictions for the 10B natural-parity level spectrum and transverse form factors, they were less successful for C2 form factors: not only is there a considerable dependence on the 1p-shell interaction, but these models give just 45% of the total observed C2 transition strength. Only a 10% improvement was realized by expanding the shell model space to include 2ħω configurations. The inclusion of even higher-excited configurations by means of core polarization calculations was essential to remove the remaining shortfall.