Glycerolipid biosynthesis by porcine adipose tissue homogenates did not yield the 90+% triacylglycerol observed in situ. Consequently, we compared intact tissue slices and various subcellular fractions to characterize the usefulness of such systems to assess glycerolipid biosynthesis in vitro. Glycerolipid biosynthesis by porcine adipose tissue homogenates was measured in vitro using either [14C]-fatty acid or [14C]-glycerol-3-phosphate (G3P) as a radiolabelled substrate. Removal of residual 14C-labelled fatty acid from lipid extracts was difficult. Because G3P is soluble in water, residual [14C]-G3P separated easily from the glycerolipid-containing organic phase and, thus, was the preferred radiolabelled substrate. With tissue slices, glycerol and G3P were minimally incorporated into lipid so that [14C]-fatty acid was the preferred radiolabelled tracer. A washing procedure followed by thin layer chromatography was devised to separate residual [14C]-fatty acid from glycerolipids, including phospholipids. Fatty acid esterification into glycerolipids in tissue slices yielded about 4% phospholipids, whereas with homogenates, esterification yielded up to 50% phospholipids. Comparison of several subcellular fractions indicated that microsomes contained most of the glycerolipid biosynthetic activity when activity was expressed on a protein basis. However, when activities were expressed on a tissue wet weight basis, the 700 x g infranate and the 10,000 x g supernate had about equal activity that was far greater than the microsomes. The 700 x g infranate was the preferred enzyme preparation for assay of the entrance of G3P into the pathway as well as the capacity to synthesize triacylglycerol. Several methods of freezing and storing tissue or 700 x g infranates were not acceptable. Freezing of the 700 x g infranate in liquid N2 with storage at -80 degrees C may be an acceptable procedure.