Photosynthesis is a biological process whose efficiency depends on the coordinated interaction of several participating protein complexes that reside in the photosynthetic membrane. Our understanding of the photosynthetic membrane structure is largely derived from studies on fractionated membranes and from conventional electron microscopy (1, 2), and our knowledge of the spatial and functional interaction between principal membrane complexes is quite limited. Such information is directly relevant to excitation energy transfer as well as to the diffusion of mobile electron carriers (3, 4) in photosynthetic membranes of oxygen-evolving plants. Major integral protein complexes include ATP synthase, cytochrome b6/f, photosystem I (PSI), and photosystem II (PSII) and LHCII in green plants. In phycobilisome (PBsome)-containing organisms-namely, red algae and cyanobacteria-the LHCII complex appears to be absent and PBsomes on the stromal surface serve as the principal light-harvesting antennae (5). In PBsome-containing organisms it is generally assumed that the arrangement of the photosystems in the thylakoid is predictable from the PSII/phycobiliprotein ratio (6, 7) and that the PBsome distribution on the stromal surface is indicative of the PSII location (8, 9). Since PBsomes with PSII activity have been isolated, a direct functional association exists in vivo (10), although a direct interaction between PBsomes and PSI cannot be ruled out. The following questions are being addressed here: (i) how many PSI and PSII centers exist per unit of surface area, (ii) is the distribution of PSI and PSII uniform throughout the membrane, (iii) are the photosystems present as monomers or as clusters, and (iv) if they are clustered, does the cluster size change under different growth conditions?By conventional immunocytochemical methods, the subcellular distribution of proteins in plants is well established, including specific localization of a number of chloroplast components in land plants and algae (1,11,12). We have shown that this approach can be used to provide a quantitative measure of the relative amounts of PSI and PSII in Porphyridium cruentum (13,14). However, since antigenic sites are accessible only on exposed surfaces of embedded sections, the absolute number of particular components is greatly underestimated, and spatial distances between individual complexes are difficult to ascertain.To ascertain the spatial relationships of PSI and PSII requires intact membranes and a means of identifying the photosystems. By freeze-fracture, 10-nm particles exposed on the exoplasmic fracture face are regarded as representing PSII centers. The isolation of PSII particles and their visualization in lipid vesicles support this attribution (15). Also, there is good evidence that some 10-to 13-nm particles on the protoplasmic fracture face represent PSI (2), but one cannot identify individual particles as photosystems. Since PSI and PSII particles are not identifiable in the same fracture-face, meaningful spatial correlations...