Measurements of the angular distribution of photoelectrons and the associated Auger electrons, together with their angular correlation, have allowed the relative amplitudes and phases of the s and d matrix elements for 4p 3/2 photoionization of Sr to be determined. The results indicate that there is destructive interference between the s and d channels and that the photoionization parameters show little dependence on energy.All the measurable quantities in an atomic photoionization process, the absolute cross section, dichroism, and the angular distributions and polarizations of emitted particles, can be predicted if the dipole matrix elements for the photoionization process are known, providing an appropriate coupling scheme has been selected. Experiments where all these parameters are extracted are called, for this reason, 'complete' photoionization experiments (Heinzmann 1980a, b). The first complete experiment, involving coincidence measurements between a photoelectron and the subsequent Auger electron, was performed on the 4d photoionization of Xe by Kämmerling and Schmidt (1993).Recently De Fanis et al (1999) reported the results of measurements of angular distributions of both the 3p 3/2 photoelectrons of atomic Ca: Ca 3p 6 4s 2 1 S 0 + hν → Ca + 3p 5 4s 2 2 P 3/2 + e −(1) and the Ca L 3 M 1 M 1 Auger electron:and of the angular correlation between them over the photon energy range hν = 36-42 eV. From these measurements the amplitude ratio and phase difference of the dipole matrix elements for the 3p → s and d channels were extracted. These results were compared with a RPAE calculation; the conclusion was that the LSJ coupling scheme for the process investigated was entirely appropriate. The work of De Fanis et al (1999) extended towards threshold the results of Lörch et al (1999), the latter covered the range hν = 45-95 eV.