Isolated maize (Zea mays var. kelvedon glory) etioplasts have been used to investigate the relationships between the spectral shifts and ultrastructural changes which occur during light-induced chloroplast development. After primary photoconversion, the Shibata shift was observed as a change from 680 to 670 nm in the chlorophyllide absorption maximum. When 1.5 nM ATP was added to the incubation medium the maximum was 675 nm even after 3.5 hours of illumination. Difference spectra for this effect indicate ATP inhibition of the Shibata shift. Two bands with maxima at 682 and 669 nm can be used to fit spectra of both ATP-treated and control etioplasts, the estimated proportions of chlorophyllide 682 being 36% and 6%, respectively. Quantitative analysis of electron micrographs of the etioplasts showed that the frequency of untransformed prolameliar bodies was also higher in the presence of ATP (73% untransformed compared to 22% in the absence of ATP). A similar correlation was observed when transformation was measured for two etioplast fractions which show the shift to different extents. These results imply that the Shibata shift and prolamellar body transformation are related events, both being inhibited by the presence of ATP. ATP may therefore have an important role in regulating the early stages of plastid development.The early stages of the light-induced transformation of greening etioplasts into chloroplasts in leaves involve several shifts in the position of the chlorophyll(ide) absorption maximum (1,8,11,18,28,(30)(31)(32) (8,12), suggesting that the changes in association between different pigment molecules or between pigment molecules and other membrane components is the cause of the shift (18). This hypothesis is supported by the demonstration that the shift also occurs in isolated Chle-holochrome suspensions (3,14,26), and other recent evidence indicates that a change in holochrome conformation occurs during the shift (14,22,27). An alternative view is that phytylation is the cause of the Shibata shift (29) but there is little direct experimental support for this view.The rate of the Shibata shift is variable and appears to depend on several parameters. Electron microscopical studies have indicated that the shift is itself correlated with specific membrane changes occurring in the developing plastid; according to Von Wettstein's group, prolamellar body dispersal can occur simultaneously with the Shibata shift in barley but they add that in other species the relationship between the spectral and ultrastructural changes is not always as clearly seen (13, 43).Studies using whole plants in which attempts have been made to examine the control of the Shibata shift and its relationship to membrane changes have been hindered by the complexity of the metabolism of the intact tissue and by the difficulty in inducing alteration of specific processes from outside the whole organism. Recently systems of isolated etioplasts have been described by Wellburn and Wellburn (40,41,42) and by Horton and Leech (15,16)