Charging of the workpiece in electron beam lithography is well recognized as a source of pattern placement error. Despite considerable previous effort there is Little quantitative understanding of the problem. A new technique was recently reported in which the surface potential of a resist film during and after exposure was measured directly. Here we describe results obtained using an improved version of the technique for charging of resists under a wide variety of conditions. Some results are as expected, e.g., thicker resists tended to charge more negatively than did thinner ones. Other results are surprising; for example, under certain conditions (7 keV, 0.4 mu m polybutene sulfone) there was zero potential at the surface and at higher energies (10 keV) the surface charged positively (0.7 V). The detailed mechanism for positive charging under these conditions is still unclear.