High spectral resolution observations of Jupiter performed at the NASA/Infrared Telescope Facility near (13.5 \mu \mathrm{m}) are presented. Two spectral ranges including the (R(7)) and (R(11)) lines of the (\nu_{2}) band of (\mathrm{HCN}) were observed at high signal-tonoise. No evidence for HCN absorption in excess of (2 %) of the continuum was found, in contrast to the report by Tokunaga et al. (1981, Icarus 48, 283-289). An upper limit on the tropospheric HCN mixing ratio of (1 \times 10^{-9}) was derived, assuming a uniform value up to the tropopause. When (\mathrm{HCN}) condensation in the upper troposphere is taken into account, only a looser upper limit around (1.2 \times 10^{-8}) can be set on the deep mixing ratio. The lack of emission cores at the position of the (\mathrm{HCN}) lines provides an upper limit of (2 \times 10^{-4} \mathrm{~cm})-amagat on the (\mathrm{HCN}) column density in the stratosphere (corresponding to a maximum average mixing ratio of (8 \times 10^{-10}) above condensation level). A critical reanalysis of Tokunaga et al.'s observations is finally presented, leading to the conclusion that the previously reported detection of HCN is questionable. O 1995 Academic Press, inc.