As I understand it, the conclusion that our spatially flat universe is expanding at an accelerating rate is supported by observations of 'standard candles' (type S1a supernovae) which appear less bright than expected when their z exceeds about 0.3 --- . Robert Kirschner has discussed these observations in his popular book The Extravagent Universe. He dismissed the possibility that the supernovae were instead dimmed appreciably by intergalactic dust. In his Fig. 10.7, for example, the dimming seems to be quite a small effect. Smoothed out, it seems to be about 0.3 magnitudes as z approaches 1, which is comparable with the observational scatter for individual quasars. These are not easy observations. They have considerable cosmological implications. In a http://arxiv.org/PS_cache/arxiv/pdf/0902/0902.4240v1.pdf" [Broken] just out Menard, Scranton, Fukugita and Richards report an analysis of SDSS data which suggest that dust may indeed affect the dimming of distant supernovae. They comment that: "The magnitude of this effect and its impact on dark energy constraints must be investigated and quantified." and later add that: "This will affect the brightness estimates of Type Ia supernovae at high redshift, which require high precision in order to maximize their constraints on cosmological parameters." The dimming they talk of is about .03 magnitudes up to z = .5 and possibly .05 to .09 at z = 1, if I read correctly. I'd appreciate an assessment of their results in relation to those of say, Perlmutter and Kirschner to date by folk more competent than I to judge their significance.