[en] Complete text of publication follows. In many cases in nuclear astrophysics, the investigated quantity is proportional to the yield of the emitted γ-radiation. At relatively small energies, which occur in a stellar environment, this yield is tiny. If a study is performed close to these energies, the knowledge of the laboratory γ-ray background is crucial. It is important to keep the background as low as possible, to reach higher sensitivity. In this work, the effect of an active shield (veto detector), and a passive shield (underground location) on the background is shown. For the present studies, a Clover detector has been used. It is equipped with a surrounding bismuth germanate (BGO) scintillator. With anticoincidence electronics, the BGO is used as an escape-suppression veto detector. It can also act as an efficient veto against muons penetrating the germanium detector volume. The underground site where the background measurements were performed is the Felsenkeller. It is a shallow underground laboratory with 45m thick rock cover (110m water equivalent). As a comparison the spectra taken from recorded at Laboratori Nazionali del Gran Sasso (LNGS) are also shown. LNGS is a deep underground laboratory with 1400m rock coverage (3400m water equivalent). For in-beam experiments, the counting rate in the high energy continuum is of paramount importance, because γ-rays from reactions with high Q-values can be emitted. In this region cosmic ray induced events dominate the laboratory background. On the Earth's surface and at the Felsenkeller, the main source of this background is the energy loss of penetrating muons in the germanium. Deep underground, because of the factor of 10⁶ reduction in the muon flux, the main background sources are neutron reactions, producing high energy gammas. In the present study, the BGO veto leads to a factor of 160 reduction in the continuum counting rates, which combined with the muon flux reduction (factor of 40) at Felsenkeller leads to a background counting rate comparable to that of LNGS. (Fig. 1.). In summary, using an active shield in a shallow underground environment, a background level comparable to that of a deep underground laboratory can be reached. Further details about the laboratory background investigations can be found in [5]. T.S. acknowledges a Herbert Quandt fellowship at Technical University Dresden.