Sands from various geographic locations reduce N2 from the air to NH3 and traces of N2H4 on exposure to sunlight. This N2 photofixation occurs under sterile conditions on the surface of finely dispersed titanium minerals such as rutile, utilizing reducing equivalents generated through the photolysis of chemisorbed H20. Abiological N2 photofixation is suggested to be part of the nitrogen cycle in arid and semiarid regions. It is estimated that about 10 x 106 tons of N2 is photoreduced on the total surface of the earth's deserts per year.The band-gap of TiO2 of 2.9-3.2 eV (290-335 kJ/mol) is larger than the energy required for water dissociation (1.23 eV), rendering TiO2 a good substratum for water splitting into H2 and 02 on irradiation with near-UV light, as was first shown by Fujishima and Honda (1,2 As outlined elsewhere (3), the reduction of N2 to NH3 and N2H4 appears to proceed in a stepwise fashion on 2-electron reducing sites at the TiO2 surface, via chemisQrbed N2H2 as the intermediate. In similar fashion, C2H2 is photoreduced to C2H4 (Eq. 4): TiO2 C2H2 + H20 + hv C2H4 + 1/2 02.[4] Because the photoreduction of C2H2 is significantly inhibited by N2, it was assumed (3) that both substrates are reduced at the same binding sites.The fact that N2 is photoreduced even in the presence of oxygen and on exposure of the Fe-doped TiO2 powders to normal sunlight (3) and the wide distribution of titanium on the earth's crust suggested that molecular N2 is fixed photochemically on sun-exposed titanium oxide containing sands or minerals without the aid of microorganisms, notably in arid or semiarid regions of the earth. The correctness of this hypothesis was first verified through studies with sand from a southern California desert in 1979 (4).In the present paper we report that results of more extensive studies with sand samples collected from locations given in Table 1.Several theories of abiological N2 fixation were formulated, and in part substantiated experimentally, more than 30 years (7), but his theories were almost universally rejected and usually are not mentioned in the contemporary literature. Because Dhar used sand from the Jumna River near Allahabad for N2 photofixation experiments in 1951 (6), we included a sample from this location in our studies. We found it to be rich in rutile and photocatalytically active under our experimental condi-tions. The sands were tested for photoreducing activity under sterile conditions on exposure to sunlight, using '5N2-enriched N2 as the substrate. Experiments were also performed with UV light as the light source with N2 as well as with C2H2 as the substrate.The sands were also tested for N2 photooxidation activity. Bickley and Vishwanathan (8) reported that H202-pretreated TiO2 powder promotes the photooxidation of N2 to a chemisorbed species yielding NO on thermolysis. The mechanism of this reaction is unknown. However, NH3 is known (9) to be photooxidized on TiO2 surfaces in the presence of 02 without requiring H202 pretreatment. Hence, under our experiment...