The shapes of three typical examples of electron distribution functions, which have been observed by Helios 2 in the solar wind, are analyzed and compared with theoretical predictions. We have considered a distribution function with a “narrow strahl” (narrow beam), which is extremely anisotropic and skewed with respect to the magnetic field direction at particle energies above 100 eV, a distribution function with a “broad strahl” (broad beam), which is less anisotropic and skewed, and finally a nearly isotropic distribution function which, however, shows a slight bidirectional anisotropy. The main results are as follows: (1) For each distribution function we may discern a “break,” i.e., a sudden change in the slope of the distribution function, separating the “core” at lower energies from the “halo” at larger energies. For the anisotropic distributions a significant break is observed in velocity directions opposite to the strahl and perpendicular to it but not along the strahl. Here the energy of the break (breakpoint energy) may be determined both by the interplanetary electrostatic potential and by collisions. In contrast, for the nearly isotropic distribution function, a significant break is observed for all velocity directions, and the breakpoint energy may be determined by collisions only. (2) The strahl observed at larger energies in the anisotropic distribution functions can be qualitatively explained by existing theoretical approaches describing the propagation of electrons in the solar wind. However, at least for the distribution function with the broad strahl as well as for the nearly isotropic distribution function, the halo electrons should be scattered by unknown anomalous scattering processes, which do not show a strong energy dependence. (3) For the anisotropic distribution functions we find a velocity shift between the peak of each distribution function and the solar wind bulk velocity, which is typically 100 km s −1 to 300 km s −1 . This shift is drastically reduced compared to the shift predicted by exospheric theory, indicating strong frictional processes between electrons and ions. However the results do not settle the question whether this friction is provided by the combined action of wave‐particle interactions and Coulomb collisions or by Coulomb collisions only. For the nearly isotropic distribution function this shift is probably not significantly different from zero. In this case it may be determined by some anomalous processes and/or trapping in closed magnetic field structures. (4) For the anisotropic distribution functions the heat flux is carried mainly by the strahl. For the nearly isotropic distribution function most of the heat flux is carried by the core electrons. For this distribution, part of the halo electrons carry heat flux in the opposite direction, and the total heat flux is probably not significantly different from zero. (5) The pitch angle distribution in the energy regime of the halo may provide some indications for the global structure of the magnetic field.