The quest to detect gaseous HCl, HBr, and HF in the atmospheres of Jupiter and Saturn has led to a tentative detection of 1 ppb HCl near Saturn's cloud deck. The detection is puzzling because, while these hydrogen halides may be present several scale heights below the clouds, they are expected to react with ammonia to form solid ammonium halide salts in the upper troposphere. I show that the loss timescale for condensation of gaseous hydrogen halides onto particles is ∼10 3–10 5 s for realistic cloud densities and particle sizes, which is much less than the ∼10 8 s residence time of upper tropospheric air. The hydrogen halides can only survive transport up to the cloud layer if less than 1 in 10 6 of their collisions with particle surfaces leads to condensation, which is unlikely. Even in the absence of foreign particles, homogeneous nucleation would probably prevent supersaturations in excess of a few hundred, which is ∼10 20–10 40 times too low to explain the observation. These calculations therefore suggest that hydrogen halides cannot exist at part-per-billion levels in the upper troposphere. The interplanetary source of halogens is also too low to produce detectable quantities of hydrogen halides except perhaps at pressures less than 1 mbar. A possible detection of chlorine by the Galileo probe at pressures exceeding 9 bars on Jupiter may be consistent with the equilibrium abundance of gaseous HCl or NH 4Cl.