The E01-011 experiment at Jefferson Laboratory (JLab) studied light-to-medium mass Lambda hypernuclei via the (e,e'K+) electroproduction reaction. Precise measurement of hypernuclear ground state masses and excitation energies provides information about the nature of hyperon-nucleon interactions. Until recently, hypernuclei were studied at accelerator facilities with intense pi+ and K- meson beams. The poor quality of these beams limited the resolution of the hypernuclear excitation energy spectra to about 1.5 MeV (FWHM). This resolution is not sufficient for resolving the rich structure observed in the excitation spectra. By using a high quality electron beam and employing a new high resolution spectrometer system, this study aims to improve the resolution to a few hundred keV with an absolute precision of about 100 keV for excitation energies. In this work the high-resolution excitation spectra of 12B-Lambda, 7He-Lambda, and 28Al-Lambda hypernuclei are presented. In an attempt to emphasize the presence of the core-excited states we introduced a novel likelihood approach to particle identification (PID) to serve as an alternative to the commonly used standard hard-cut PID. The new method resulted in almost identical missing mass spectra as obtained by the standard approach. An energy resolution of approximately 400-500 keV (FWHM) has been achieved, an unprecedented value in hypernuclear reaction spectroscopy. For 12B-Lambda the core-excited configuration has been clearly observed with significant statistics. The embedded Lambda hyperon increases the excitation energies of the 11B nuclear core by 0.5-1 MeV. The 7He-Lambda spectrum has been observed with significant statistics for the first time. The ground state is bound deeper by roughly 400 keV than currently predicted by theory. Indication for the core-excited doublet, which is unbound in the core itself, is observed. The measurement of 28Al-Lambda provides the first study of a d-shell hypernucleus with sub-MeV resolution. Discrepancies of up to 2 MeV between measured and theoretically predicted binding energies are found. Similar disagreement exists when comparing to the 28Si-Lambda mirror hypernucleus. Also the core-excited structure observed between the major s-, p- and d-shell Lambda orbits is not consistent with the available theoretical calculations. In conclusion, the discrepancies found in this study will provide valuable input for the further development of theoretical models.