Theoretical calculations on the photodissociation of water in the first absorption band have been used to test the accuracy of three available potential energy surfaces for the first excited state of water: the well-known coupled electron pair approximation potential of Staemmler and Palma [Chem. Phys. 93, 63 (1985)], and two new multireference double excitation configuration interaction surfaces: the Dobbyn-Knowles surface (unpublished), and the Leiden surface [R. van Harrevelt and M. C. van Hemert, J. Chem. Phys. 112, 5777 (2000)]. Exact quantum mechanical calculations, using the wave packet approach, have been performed for J">0, where J" is the initial rotational state of the water molecule. The cross section was found not to depend strongly on the rotational state, so that it is reasonable to compare calculated cross sections for J"=0 with experimental room temperature cross sections. Small and simple corrections were applied to the potential energy surface to improve the agreement between theory and experiment for the cross section of H2O. Spectra for D2O and vibrationally excited water molecules calculated with all three corrected potential energy surfaces were in good agreement with experiments. A comparison between calculated OH(X) or OD(X) vibrational distributions, and recent kinetic energy release measurements of the H or D atoms produced in the 157.6 nm photodissociation of water and its isotopomers [Yang , J. Chem. Phys. 113, 10597 (2000)], however, suggests that the Leiden surface is more accurate than the two other surfaces.