This minireview in memory of Daniel I. Arnon, pioneer in photosynthesis research, concerns properties of the first and still only known alternative photophosphorylation system, with respect to the primary phosphorylated end product formed. The alternative to adenosine triphosphate (ATP), inorganic pyrophosphate (PPi), was produced in light, in chromatophores from the photosynthetic bacterium Rhodospirillum rubrum, when no adenosine diphosphate (ADP) had been added to the reaction mixture (Baltscheffsky H et al. (1966) Science 153: 1120-1122). This production of PPi and its capability to drive energy requiring reactions depend on the activity of a membrane bound inorganic pyrophosphatase (PPase) (Baltscheffsky M et al. (1966) Brookhaven Symposia in Biology, No. 19, pp 246-253); (Baltscheffsky M (1967) Nature 216: 241-243), which pumps protons (Moyle J et al. (1972) FEBS Lett 23: 233-236). Both enzyme and substrate in the PPase (PPi synthase) are much less complex than in the case of the corresponding adenosine triphosphatase (ATPase, ATP synthase). Whereas an artificially induced proton gradient alone can drive the synthesis of PPi, both a proton gradient and a membrane potential are required for obtaining ATP. The photobacterial, integrally membrane bound PPi synthase shows immunological cross reaction with membrane bound PPases from plant vacuoles (Nore BF et al. (1991) Biochem Biophys Res Commun 181: 962-967). With antibodies against the purified PPi synthase clones of its gene have been obtained and are currently being sequenced. Further structural information about the PPi synthase may serve to elucidate also fundamental mechanisms of electron transport coupled phosphorylation. The existence of the PPi synthase is in line with the assumption that PPi may have preceded ATP as energy carrier between energy yielding and energy requiring reactions.