Expressions for analyzing small-angle scattering data from semidilute solutions of polymers in a good solvent over a broad range of scattering vectors are examined. Three different scattering function expressions are derived from Monte Carlo simulations. The expressions are similar to those of polymer reference interaction site models, with a scattering-vector-dependent direct correlation function. In the most advanced model, the screening of excluded-volume interactions beyond the overlap concentration is taken into account. Two simpler expressions, in which the screening of excluded-volume interactions is not included, are also applied. The three models are tested against small-angle neutron scattering (SANS) experiments on polystyrene in deuterated toluene for a broad range of molar masses and concentrations over a wide range of scattering vectors. For each model, simultaneous fits to all the measured scattering data are performed. The most advanced model excellently reproduces the SANS data over the full range of the parameters. The two simpler models fit the data almost equally well. On the basis of an extensive study, an optimal fitting strategy can be recommended for experimentalists, who want to analyze small-angle scattering data from polymers at any concentration. For data sets that do not contain data on the single-chain scattering function, the simpler model is recommended; it uses a direct correlation function equal to the form factor of an infinitely thin rod, which is independent of the concentration and molar mass. (C) 2004 Wiley Periodicals, Inc.