Properties of the first excited state of the nucleus 9Be are discussed based on recent (e,e) and (gamma,n) experiments. The parameters of an R-matrix analysis of different data sets are consistent with a resonance rather than a virtual state predicted by some model calculations. The energy and the width of the resonance are deduced. Their values are rather similar for all data sets, and the energy proves to be negative. It is argued that the disagreement between the extracted B(E1) values may stem from different ways of integration of the resonance. If corrected, fair agreement between the (e,e) and one of the (gamma,n) data sets is found. A recent (gamma,n) experiment at the HIgS facility exhibits larger cross sections close to the neutron threshold which remain to be explained.