The activity of adenosine 5' triphosphate sulfurylase was determined in crabgrass mesophyll cells, bundle sheath strands, and whole leaf extracts. The enzyme was assayed by following molybdate-dependent pyrophosphate release from ATP, 35So42-incorporation into adenosine 5' phosphosulfate, and ATP synthesis dependent upon adenosine 5' phosphosulfate and inorganic pyrophosphate. With all assays, greater than 90% of the activity was found in extracts from bundle sheath strands. The activities in whole leaf extracts were consistently intermediate between the activities of mesophyll and bundle sheath extracts and extract-mixing experiments gave no indication of enzyme activation or inhibition in vitro. Whole leaf activities were several hundred-fold less than concurrent measurements of ribulose 1,5-bisphosphate and phosphoenolpyruvate carboxylase activities, which is interpreted as being consistent with the relative amounts of elemental carbon and sulfur found in higher plants. A hypothesis is presented for the intercellular compartmentation of sulfur assinilation in relationship to N03-and CO2 assimilation in leaves of C4 plants.Higher plants assimilate the essential elements carbon, nitrogen, and sulfur, primarily as the oxides C02, NO3-, and s042-. The transformations of these oxides into organic molecules are coupled to light-dependent reduction and ATP synthesis reactions in leaves. In plants with the C4 pathway of CO2 assimilation, the transformation of CO2 and NO3 is compartmentalized into specific cell types. While atmospheric CO2 initially is fixed into organic acids in the mesophyll cells of C4 plants, the carboxylation of CO2 to form 3-PGA3 occurs exclusively in the bundle sheath cells (6). Conversely N03-, which is made available to the chlorenchyma cells of plants by way of the transpiration stream, moves past the bundle sheath cells of C4 plants and is reduced exclusively in the mesophyll cells (7,17). The cellular compartmentation of carbon-and nitrogen-assimilating enzymes is a primary factor contributing to high rates of carbon assimilation (6) and nitrogen use efficiency (7,9,17) . In some instances the tissue slices were preincubated in 1% pectinase and 1% cellulysin to increase the yield of cell types. The procedure followed was that described by Chen et al. (11) In a control to correct for nonspecific ATPase activity, NaCl was used in place of Na2MoO4. After incubation for timed periods at 37 C, the reaction was stopped-by the addition of 0.5 ml of ice cold 15% trichloroacetic acid. The reaction tubes were transferred to ice. The tubes were centrifuged at 3,000g for 10 min at 4 C and an aliquot (generally 0.4 ml) was removed for Pi determination.