Adipose tissue-derived tumor necrosis factor (AT-TNF) protein and messenger RNA (mRNA) has been shown to correlate with insulin resistance in some studies. However, in a study using different aged Fischer 344 rats, AT-TNF activity correlated more strongly with cell size than with fasting plasma insulin. The present study was undertaken to more carefully examine the relationship among AT-TNF, adipose cell size, and insulin action using more precise measures of insulin action. Basal and hyperinsulinemic, euglycemic clamps were performed in male Sprague Dawley rats at four different ages (8, 13, 21, and 61 weeks old). [3-(3)H]glucose and 2-deoxy-D-[1-(14)C]glucose were used to assess glucose kinetics and tissue-specific glucose uptake. Because TNF activity represents the summation of TNF synthesis, secretion, and the amount of soluble inhibitors present, TNF activity was measured using a bioassay, in addition to measuring TNF protein and mRNA levels. AT-TNF activity increased significantly with age, as did the glucose infusion rate, a measure of whole body insulin resistance. However, AT-TNF activity did not correlate with any parameter of insulin action measured during the hyperinsulinemic, euglycemic clamps. In epididymal fat, AT-TNF activity correlated with: glucose infusion rate: r = -0.50, P = 0.17; rate of appearance: r = -0.19, P = 0.35; rate of disappearance: r = 0.08, P = 0.69. As was noted before, AT-TNF activity correlated well with fat cell size (r = 0.76, P < 0.001 in epididymal fat; r = 0.58, P = 0.007 in SUB fat). These data suggest that although AT-TNF activity and insulin resistance increase with age, the two are not functionally related. These data do not eliminate the potential role of nonadipose TNF in the regulation of insulin action.