Current research suggests that AIDS has been in human circulation for at least 100 years and that its transformation in to a pandemic was triggered by the European colonization of the African continent.
The Origin of HIV
Scientists set the origins of the AIDS pandemic to the first recorded case in 1976, but without historical samples it was impossible to trace back the origins of the virus. The closest researchers had come to revealing the history of the virus was a study by Keele et al. in 2006 which linked the HIV-1 virus to a "closely related" simian immunodeficiency virus (SIV) discovered in captive chimpanzees of the subspecies Pan troglodytes troglodytes originating in southeastern Cameroon.1
Until recently the only known sample of HIV-1 from before 1976 was from blood extracted from a tissue sample obtained in Léopoldville (now Kinshasa, Democratic Republic of the Congo) and preserved in paraffin wax. The tissue sample (identified as ZR59) was taken from a "Bantu male" in 1959. The discovery of HIV-1 in this historical sample was reported by a research team in Nature in 1998.2
Further research, reported by Korber et al in 2000 suggested that HIV-1 in humans originated between 1915 and 1941, with an optimum estimate for the "common ancestor of the main HIV-1 group to be 1931.3
How Korber's Estimate was Obtained
All viruses mutate over time and research into HIV had revealed a surprisingly large degree of mutation for something apparently dating back to the mid 70s. By comparing the genetic make up of HIV from different samples it is possible to predict when a "common ancestor" of the virus existed. By comparing the genetic differences between modern HIV-1 samples and that extracted from the 1959 sample, Korber and his colleagues were able to estimate an origin date of 1931. (The range of potential dates given as 1915 to 1941 represents the uncertainty of the analysis which is based on a single historical sample.)
Further Research
As reported in October 2008, David Worobey (of the Ecology and Evolutionary Biology department, University of Arizona in Tucson) and his colleagues recovered a new sample (identified as DR60), once again from Léopoldville (Kinshasa, DRC). 4 The sample this time was taken in 1960, and from a female. The team first had to deal with the genetic damage which had occurred when the tissue (a lymph node biopsy) had been drenched in chemicals, which broke up the genome, and then preserved in paraffin wax. DNA and RNA sequences were extracted, a process which was repeated by a second team at Northwestern University in Chicago, Illinois.
Worobey claims to have completed "the first comparative evolutionary genetic study of early pre-AIDS epidemic HIV-1 group M viruses."4 Genetic differences, due to mutation, of the order of 12% between HIV recovered from the two historic samples, DRC60 and ZR59, suggest that HIV had been present in humans for a significant period before the 60s. The same genetic analysis suggests they derived from the same ancestral SIV chimp virus.
"The sizeable genetic distance between DRC60 and ZR59 directly demonstrates that diversification of HIV-1 in west-central Africa occurred long before the recognized AIDS pandemic."4
Further analysis, comparing over 100 modern HIV samples to the two historical examples allowed the researchers to create a genetic "family tree" which goes back around 100 years to a common ancestor.
"Our best estimate for when HIV entered humans is 1908, but it could have been from 1884 to 1924".4
It is believed that HIV repeatedly crossed from chimps to humans prior to 1908, but never spread because those infected were members of scattered and isolated communities. In order for the virus to act as a epidemic, it needed contact between people, the sort of contact which you get in urban areas.
By coincidence, 1908 is around the time that major cities were growing in central Africa - Kinshasa, Douala, Brazzaville, Yaounde, and Bangui were founded by European colonists between 1880 and 1900, and by 1910 the populations of these cities were booming.
Worobey states "high risk [sexual] behaviors found in [these cities] are necessary to allow one of these sporadic viral jumps to get a toehold in the human population"4 but that the good news is that since it failed to gain such a toehold prior to urban living "HIV lives right on the edge of extinction"4. Worobey suggests that modern developments (such as education, testing, prevention campaigns, and antiretroviral drugs) could lead to the end of the epidemic even without the creation of a vaccine.
References:
1. Keele, B.F. et al., "Chimpanzee Reservoirs of Pandemic and Nonpandemic HIV-1", Science 313, pp 523-526, 28 July 2006, DOI: 10.1126/science.1126531.
2. Zhu, T. et al., "An African HIV-1 sequence from 1959 and implications for the origin of the epidemic", Nature 391, pp 594-597, 5 February 1998, DOI:10.1038/nature35400.
3. Korber, B. et al., "Timing the Ancestor of the HIV-1 Pandemic Strains", Science 288, pp 1789-1796, 9 June 2000, DOI:10.1126/science.288.5472.1789.
4. Worobey, M. et al., "Direct evidence of extensive diversity of HIV-1 in Kinshasa by 1960", Nature 455, pp 661-664, 2 October 2008, DOI:10.1038/nature07390.
Sources:
Debora MacKenzie, "First cities in central Africa triggered AIDS", New Scientist, 4 October 2008, p 10.
Heidi Ledford, "Tissue sample suggests HIV has been infecting humans for a century", Nature, published online, 1 October 2008.

