The total infrared (IR) luminosity is very useful for estimating the star formation rate (SFR) of galaxies, but converting the IR luminosity into an SFR relies on assumptions that do not hold for all galaxies. We test the effectiveness of the IR luminosity as an SFR indicator by applying it to synthetic spectral energy distributions generated from three-dimensional hydrodynamical simulations of isolated disc galaxies and galaxy mergers. In general, the SFR inferred from the IR luminosity agrees well with the true instantaneous SFR of the simulated galaxies. However, for the major mergers in which a strong starburst is induced, the SFR inferred from the IR luminosity can overestimate the instantaneous SFR during the post-starburst phase by greater than two orders of magnitude. Even though the instantaneous SFR decreases rapidly after the starburst, the stars that were formed in the starburst can remain dust-obscured and thus produce significant IR luminosity. Consequently, use of the IR luminosity as an SFR indicator may cause one to conclude that post-starburst galaxies are still star forming, whereas in reality, star formation was recently quenched.
