Dosimetric characteristics of a number of clinically applied electron beams were analyzed as part of a dosimetry intercomparison program performed at the radiotherapy centers in The Netherlands. Absorbed dose values, determined under reference conditions, were compared during site visits with stated dose values. The mean deviation was 0.2% with a standard deviation (SD) of 2.5%. The maximum deviation was 5.7%. The largest differences were due to differences in the calibration procedures and differences in the numerical values of conversion factors adopted from different dosimetry protocols. In addition, a number of clinically relevant parameters of the dose distribution on the central beam axis were analyzed including the depth of the 85% and 50% relative dose, the dose reduction at the depth of the 50% relative dose and the surface dose. The average difference between the stated and measured therapeutic depth (85% dose level) was -0.4 mm with an SD = 1.2 mm. Due to this dosimetric uncertainty observed a safety margin of about 3 mm at the therapeutic depth is recommended. The maximum difference between stated and observed mean energy of the electron beams had only a small influence, < 1%, on the absorbed dose determination. The normalized dose gradient is not an adequate parameter to describe the dose reduction beyond the therapeutic depth. The depth of a low dose level is a better parameter. The relative dose at the surface showed differences up to 10% between scanning electron beams and beams from accelerators with a single scattering foil and closed wall collimating system.