A safe and effective HIV vaccine to prevent infection and/or to moderate disease is urgently needed. Research progress has been slower than anticipated for a variety of reasons including uncertainty over which immunogen to use (i.e. recombinant subunit envelope proteins or whole HIV-1 products), confusion on which immunological markers best correlate with protection, the relevance of the HIV-1 chimpanzee model to infection in humans and the significance of the rapid evolution of HIV-1, with different clades of the virus emerging in different parts of the world. However, what some would interpret as encouraging results, from Phase I and II trials of recombinant envelope glycoprotein vaccines, have raised the question of whether the time is right to start Phase III trials in humans with immunogens that may have low to moderate efficacy. By using mathematical models and data from epidemiological studies, we examine the potential impact of such vaccines within heterosexual communities with high rates of infection. Analyses suggest that it will be difficult to block HIV-1 transmission even with very high levels of mass vaccination. The cost of sustaining high levels of herd immunity with a vaccine of short protection duration is likely to be high. However, assessments of impact over the long duration of an HIV-1 epidemic indicate that many cases of HIV infection and associated mortality can be prevented by immunogens with efficacy of 50% or less and a five year protection duration. These analyses add some support to the view that proceeding with Phase III efficacy trials may be appropriate in high HIV transmission regions even if the consensus opinion on potential efficacy of the immunogen is that it will be low.