In the “hydrogen transportation economy” personal vehicles will be powered by either fuel cells or internal combustion engines fueled by hydrogen.  This economy will provide society many benefits in terms of fuel stock diversification and reduction or elimination of CO₂ and other emissions.  Aside from the technical issues involved in making such an economy commercially viable, though, there remains the “chicken and egg” problem of building a hydrogen supply and distribution infrastructure in the absence of an existing market, and conversely inducing consumers to invest in hydrogen vehicles when the hydrogen supply structure is sparse.  This problem is both unlike and more daunting than those facing previous introductions of new vehicle technologies such as hybrid-electric vehicles or “flex-fuel” ethanol vehicles.  Thus it is of importance to understand the details of how such a transition might occur.  Under a DOE-sponsored program, we have developed a computer simulation of such a transition in a particular locality, the Los Angeles metropolitan area, using the technique of agent-based modeling.  This method is of particular utility in a case such as ours in which a large number of diverse participants interact with one another in possibly complex ways.  In our simulation, agents are of two types:  potential buyers of hydrogen vehicles and potential investors in hydrogen fueling stations.  The driver agents live and work in different locations, have different incomes and different attitudes towards new technology, and participate in different social groups.  Investor agents are characterized in terms of different levels of risk tolerance and knowledge about driver purchase intentions.  At the beginning of the simulation, almost all driver agents own conventional, petroleum-based vehicles and only a few “seed” hydrogen fueling stations exist.  Driver agents may be induced to trade in their conventional vehicles for hydrogen-powered ones by a variety of factors:  lower operating costs, the desire to be “green”, social pressures from their peer groups, etc., but may be dissuaded from doing so by noticing a lack of convenient hydrogen fueling stations as they drive around the LA area.  The balance of these opposing factors determines for each individual driver agent what type of vehicle he or she buys.  Investor agents attempt to predict the growth of hydrogen vehicles based on possibly imperfect knowledge of driver agent characteristics and intentions and base their investment decisions on these predictions.  If they predict inaccurately they suffer financially, but they are able to learn from their mistakes and refine their prediction ability.  Depending upon how rapidly investors build new infrastructure and how drivers react to the sparse but growing number of hydrogen fueling stations, the transition occurs more or less rapidly, or may fail altogether.  We will describe model results showing how such transitions occur under a number of different scenarios, and discuss what factors appear to be most important in influencing their development.