We obtained time-resolved, near-infrared spectra of Io during the 60–90 min following its reappearance from eclipse by Jupiter on five occasions in 2004. The purpose was to search for spectral changes, particularly in the well-known SO 2 frost absorption bands, that would indicate surface–atmosphere exchange of gaseous SO 2 induced by temperature changes during eclipse. These observations were a follow-on to eclipse spectroscopy observations in which Bellucci et al. [Bellucci et al., 2004. Icarus 172, 141–148] reported significant changes in the strengths of two strong SO 2 bands in data acquired with the VIMS instrument aboard the Cassini spacecraft. One of the bands (4.07 μm [ ν 1 + ν 3]) observed by Bellucci et al. is visible from ground-based observatories and is included in our data. We detected no changes in Io’s spectrum at any of the five observed events during the approximately 60–90 min during which spectra were obtained following Io’s emergence from Jupiter’s shadow. The areas of the three strongest SO 2 bands in the region 3.5–4.15 μm were measured for each spectrum; the variation of the band areas with time does not exceed that which can be explained by the Io’s few degrees of axial rotation during the intervals of observation, and in no case does the change in band strength approach that seen in the Cassini VIMS data. Our data are of sufficient quality and resolution to show the weak 2.198 μm (4549.6 cm −1) 4 ν 1 band of SO 2 frost on Io for what we believe is the first time. At one of the events (June 22, 2004), we began the acquisition of spectra ∼6 min before Io reappeared from Jupiter’s shadow, during which time it was detected through its own thermal emission. No SO 2 bands were superimposed on the purely thermal spectrum on this occasion, suggesting that the upper limit to condensed SO 2 in the vertical column above Io’s surface was ∼4 × 10 −5 g cm −2.