We present 30 optical spectra and 49 photometric epochs sampling the first 517 days after discovery of supernova (SN) 1999em and derive its distance through the expanding photosphere method (EPM). SN 1999em is shown to be a Type II-plateau (II-P) event, with a photometric plateau lasting until about 100 days after explosion. We identify the dominant ions responsible for most of the absorption features seen in the optical portion of the spectrum during the plateau phase. Using the weakest unblended absorption features to estimate photospheric velocity, we find the distance to SN 1999em to be D = 8.2 ± 0.6 Mpc, with an explosion date of HJD 2,451,475.6 ± 1.4, or 5.3 ± 1.4 days before discovery. We show that this distance estimate is about 10% closer than the distance that results if the strong Fe II λλ4924, 5018, 5169 absorption features, which have often been used in previous EPM studies, are used to estimate photospheric velocity. We examine potential sources of systematic error in EPM-derived distances and find the most significant to result from uncertainty in the theoretical modeling of the flux distribution emitted by the SN photosphere (i.e., the "flux dilution factor"). We compare previously derived EPM distances to eight SNe II in galaxies (or members of the same group) for which a recently revised Cepheid distance exists from the Hubble Space Telescope Key Project and find DCepheids/DEPM = 0.87 ± 0.06 (statistical); eliminating the three SNe II distances for which a Cepheid distance exists only to a group member, and not the host galaxy itself, yields DCepheids/DEPM = 0.96 ± 0.09. Additional direct comparisons, especially to spectroscopically and photometrically normal SNe II-P, will certainly help to produce a more robust comparison. Finally, we investigate the possible use of SNe II-P as standard candles and find that for eight photometrically confirmed SNe II-P with previously derived EPM distances and SN 1999em, the mean plateau absolute brightness is M̄v(plateau) = -16.4 ± 0.6 mag, implying that distances good to ∼30% (1 σ) may be possible without the need for a complete EPM analysis. At M̄v(plateau) = -15.9 ± 0.2 mag, SN 1999em is somewhat fainter than the average SN II-P. The general consistency of absolute SNe II-P brightness during the plateau suggests that the standard candle assumption may allow SNe II-P to be viable cosmological beacons at Z > 2.