Rural areas of the developing world are the last frontier of the information technology revolution. Telephone and internet penetration there remains a small fraction of what it is in the developed world. Limited means of electronic communication with the outside world are just one source of isolation of rural communities and economies from the forces of national and global integration, albeit an important one. Without roads and electricity, the benefits of extending ICT access would be greatly diminished. Conversely, where these other elements of infrastructure are in place, those benefits can be multiplied. The costs of ICT provision to rural areas tend to be higher than to more densely populated urban areas, and the ability to pay of potential subscribers lower. In recent years, a number of interesting experiments has been initiated to extend low-cost telephone and, in some cases, internet access to low-income rural communities. This paper reviews some of these, with a particular emphasis on whether they are likely to prove financially sustainable. Surveys of rural households' willingness to pay for telephone service point to its potential commercial viability. Franchise models of shared-access provision would appear to have the most favourable economics. Through standardisation and demand aggregation, they offer the prospect of reaping economies of scale in hardware and software procurement as well as in provision of technical support, and enhanced bargaining power in negotiating interconnection fees and leased line prices. Such models provide opportunities to small entrepreneurs, at the same time creating incentives to both cost containment and rigorous financial management. If innovation is not to be stymied, however, the franchise model needs to allow for flexible local variation in implementation. While rural economies may be less complex technologically than the highly urbanised economies of the developed world, the value of timely information can be just as high -- if not higher -- in relative terms. Besides the strictly economic benefits, there can be important social benefits of maintaining long-distance contact with family members working abroad or in the city. The experience of Bangladeshi women who make up the majority of village phone operators for the Grameen network suggests that social status can be enhanced by virtue of control over a valuable resource -- information access. Governments need to consider the difficult tradeoffs among competing uses of scarce tax revenues and borrowed funds. Fortunately, in the case of telecoms and the internet, there is not the same need for substantial public investment that there is in rural roads and electricity. An adequate regulatory framework and well-designed contractual bidding process for supplying less profitable segments of the market should go a long way towards attracting private investment, including in underserved areas. In the end, this is likely to be the most direct path towards the ultimate goal of "universal service". (A bibliography is included. Transmission Option and Energy Solution information is annexed. Contains 3 footnotes, 6 boxes and 4 tables.)