There are three methods for calculating the gas flow in thin, rectangular channels anywhere in the viscous flow region (incompressible viscous flow, or compressible viscous flow with choked or unchoked exit). Two of these models have been compared to experimental data in a limited range of channel length-to-thickness ratio L/e. However, they have not been compared to experimental data over a wide enough L/e range to establish bounds on their usefulness. We have compared methods of Sasaki and Yasunaga and Kieser and Grundner and observed the former technique to agree with experimental data for the range L/e > 150 for viscous flow, unchoked exit, while the latter is in agreement over the range 8 < L/e < 400, for molecular, transition, and viscous flow (choked or unchoked exit). We examine a new model for flow in choked ducts proposed by Santeler and show that it agrees with measurements in the viscous flow region (thickness-to-mean free path ratio e/λ> 100) with choked exit for all values of L/e except for the largest. A correction to the method of Kieser and Grundner is offered which improves its accuracy for long channels in the high pressure region. © 1987, American Vacuum Society. All rights reserved.