Experimentally determined intensity correlation functions of a single-mode dye laser are compared with theoretically predicted forms given in a recent paper by Graham, Hohnerbach, and Schenzle. Although good agreement is obtained in some cases for which the parameters are chosen for best fit, large discrepancies appear for other working points of the laser. These results suggest that the dynamical theory of the dye laser needs to be modified.PACS numbers: 42.55.Mv It is well known from several experiments 1 ' 2 that the behavior of a single-mode dye laser in the region of threshold is not well described by the usual single-mode laser theory. Although the theory has been modified to include possible contributions from triplet states of the dye molecules, 3 ' 4 it does not appear that triplet states alone can adequately account for the observed switching behavior of the laser near threshold. We have suggested 2 that small pumping fluctuations, possibly connected with the flow of the dye or with the ion laser that pumps the dye laser, may be responsible for the observed effects. Measurements of the relative intensity fluctuations ((A/) 2 )/^) 2 as a function of ) were found to be reasonably consistent with such a hypothesis. 2 More recently a similar idea has been further developed into a dynamical theory by Graham, Hohnerbach, and Schenzle 5 ' 6 and Schenzle and Brand, 7 in which the fluctuations of the pump parameter are included in the equation of motion as a form of multiplicative noise. The intrinsic quantum or spontaneous emission fluctuations, on the other hand, are assumed to be comparatively unimportant above threshold.The theory of Graham, Hohnerbach, and Schenzle 5 and Schenzle and Brand 7 gives not only quite a good account of the observed instantaneous fluctuations of the laser field, but it also allows the time development of the intensity fluctuations to be tested against experiment. By choosing parameters for best fit with the data, Graham, Hohnerbach, and Schenzle succeeded in obtaining excellent agreement between one of the measured and one of the predicted two-time intensity correlation functions, 5 which suggests that the theory is on the right track. However, one curve is not decisive, because the theory contains a free parameter; one really should look at a family of intensity correlation functions at different working points of the laser for a more searching test of the model. When such a test is carried out, quite large discrepancies between theory and experiment are encountered, which suggests that the dynamical aspects of the laser model need to be modified.Graham, Hohnerbach, and Schenzle 5 and Schenzle and Brand 7 describe the laser on resonance by a dimensionless equation of motion of the usual form,