A method was developed for repeated measurements of glucose transport in human fat cells obtained by needle biopsy aspiration. Assay conditions, reproducibility and normal values of the measurements are described. Transport rates were measured in the absence and presence (25, 50, 100, 200, 800 pmol/l; 8, 80 nmol/l) of insulin using U-(14C)-D-glucose as the tracer. The extracellular glucose concentration was 1.5 μmol/l. The reproducibility of glucose transport measurements was assessed by taking two needle biopsies from opposite sides of the same subject (n =11). The mean coefficient of variation for maximal glucose transport was 11±6%. In 14 subjects, a needle biopsy sample was aspirated immediately prior to surgical removal of fat. The maximal insulin-stimulated glucose transport rates averaged 143±15 and 443±15 fl/cell · s, and the ED 50: s 218±124 and 160±28 pmol/l (NS) in fat cells prepared from needle biopsy and surgically removed adipose tissue respectively. The mean coefficient of variation for maximal glucose transport in needle vs. surgical samples was 11±2%. In 6 subjects, a surgical biopsy was taken twice, with a 1-week interval. The coefficient of variation averaged 9±2%. We conclude that measurement of glucose transport rates can be done with similar accuracy using fat cells isolated from needle biopsy aspirates and surgically removed adipose tissue. Use of needle biopsy samples permits, however, study of glucose transport in repeat samples of human fat cells, and may therefore be a useful tool for any perturbation studies.