Each year I ask the first-year chemistry students to dream about five biochemical reactions, that if industrially scaled, would change our lives. In my dreams, foremost among these would be the splitting of water to form hydrogen. A close second would be efficient, low-temperature fixation of nitrogen. Despite the fact that we know the basic structures of the proteins and prosthetic groups that mediate these microbially-derived processes, we do not understand how photosystem II or nitrogenase actually catalyse their respective reactions. The core issue, how organisms came to appropriate prosthetic groups that incorporate specific metals in protein scaffolds to direct highly organized redox chemistry, remains one of the greatest unsolved problems in biology. Despite the extraordinary explosion of knowledge of whole organism genome sequences and protein structures over the past decade, there has been surprisingly little understanding of the processes that selected the redox reactions which constitute the basic metabolic backbone of microbial life. Indeed, all the major metabolic pathways evolved in the first half Earth's history and have remained virtually unchanged over the second half, yet humans cannot replicate some of the most fundamental processes upon which our very existence is dependent. Whereas genome sequence analyses and molecular phylogenetic models provide clues to the origin of these metabolic processes, recreation of these processes de novo has thus far remained elusive. Our economic dependence on fossil organic carbon for energy is so profound that wars are fought for supply of the substrates. Clearly, if an efficient, cost-effective hydrogen generating reaction were available, an alternative, environmentally sound, globally accessible, renewable energy source would be used for transportation, electrical generation, and other forms of energy. Hydrogen supplied from the decarbonization of fossil fuels does not meet these criteria. The production of hydrogen from anaerobic fermentation of organic substrates potentially can be developed, but to be scaled to global energy demands, the production of the substrates would require vast areas of agricultural production – there simply is not enough freshwater and land area to provide the substrates for the energy demands of the future (Caldeira et al., 2003). However, the supply of hydrogen from photocatalytic water splitting would meet all criteria and is potentially achievable with artificially constructed reaction centres derived from mutant cyanobacteria. Similarly, the Haber process, which permits us to add fixed inorganic nitrogen to sustain our food supply, was developed over 80 years ago and has not fundamentally changed since. The basic reaction, in which N2 is reduced to NH3 by the addition of H2 in the presence of a catalyst at high temperature, exceeds present natural nitrogen fixation by a factor of two – and will grow as human population increases (Falkowski et al., 2000). Although there have been many efforts to incorporate nif genes into crop plants, none has yet succeeded. Were this to happen, the oversupply of fixed nitrogen to the environment, and subsequent eutrophication of coastal waters, lakes, and aquifers, would cease, while plants that provide food and fibre would have a supply of nitrogen at grain filling time – rather than an overabundance of nitrogen when seeds are germinating but a deficiency later in the growth cycle. By the end of this century, human societies will come to rely on genetically engineered microbes for hydrogen production and nitrogen fixation. In my crystal ball, mimicry of photochemical water splitting will be achieved by developing hydrogen farms based on the artificial synthesis of photosystem II reaction centres. The hydrogen farms will be coupled to nitrogen fixing protein factories, in which de novo synthesis of ammonium from N2 will be produced from the anaerobic oxidation of organic waste products which will be supplied by waste recycling streams. These industrially scaled microbial reactions will harness two basic biochemical reactions and in so doing will help meet the demands of sustainable development. Clearly the potential to mimic, alter, and improve the microbial processes of oxygenic photosynthesis and nitrogen fixation has been a dream of many researchers over the past century – but our aggregate investment in these processes remains woefully inadequate relative to the potential payback.