Four experiments were made with adult, Single Comb White Leghorn cockerels to investigate aspects of the bioassay for true metabolizable energy adjusted to zero nitrogen balance (TMEn). Collection of excreta in plastic bags held in place by harnesses (H) doubled the potential bird capacity of a standard cage arrangement relative to that wherein excreta are collected on trays (T). Housing H-birds in adjacent cages had no effect on fecal (F) and urinary (U) energy adjusted to zero nitrogen balance (FEn + UEn) irrespective of whether neighboring birds were fed or fasted. Extension of the preassay fast to 48 h did not prevent the reduction in FEn + UEn, seen in previous experiments, of H-birds relative to that of T-birds. Extension of the preassay fast to 48 h, provision of supplemental glucose, and the variation of test material inputs from 30 to 50 g did not substantially effect TMEn estimates or their precision: supplemental glucose reduced body weight loss, but the appearance of glucose in an excreta sample was a cause for concern. Provision of supplemental water by intubation during fasting and excreta-collection periods had little effect on the variance of FEn + UEn values of fed or fasted birds, and no effect on TMEn estimates. Although excreta collection with H has several advantages, there can be a high incidence of sample loss due to H slippage: 21, 2, 0, and 22% in Experiments 1, 2, 3, and 4, respectively. The reduced FEn + UEn of H-birds is a problem requiring explanation. In terms of data quality, none of the assay modifications was beneficial.