Accurate synthetic telluric spectra are required for efficient use of telescope time, in particular, with large telescopes and high-resolution NIR spectroscopy: (i) In the preparation of observations, are the telluric features at the same wavelength as spectroscopic features of scientific interest? Since water vapor is the molecule whose abundance varies most in the atmosphere, what values of precipitable water vapor are suitable to carry out successful observations? Are the observations of a telluric star required? Or better, can telluric features in the science spectrum be accurately represented by an appropriate synthetic spectrum? This point is also very important in the planning of telescope time, as observations of a telluric star may sometimes take longer than the one of the science target. (ii) In the analysis of the observations, how do telluric lines affect the scientifically interesting features in the observed spectrum? Is it possible to recover the useful information when telluric star observations could not be obtained, do not have sufficient SNR, or suffer from a significant change in instrumental or observing conditions?