Total-tract digestibility (TTD) data from 3 published studies were used to evaluate NRC (2001) estimates of metabolizable energy (ME) in calf starters (CS). Holstein bull calves (2–3 d of age at initiation of each study) were fed varying amounts of milk replacer (MR) with texturized or pelleted CS and water for ad libitum consumption. Calves (n = 60) were housed in individual pens for 8 wk (Trials 1, 2a, 3a). In 2 trials, calves were subsequently moved to group pens (4 calves/pen; n = 24) for an additional 8 wk (Trials 2b, 3b). Apparent TTD was measured at 3, 6 and 8 wk (Trial 1); 5 and 8 wk (Trial 2a); 11, 13 and 16 wk (Trial 2b); 5, 7 and 8 (Trial 3a); and 11 and 15 wk (Trial 3b). Feed and feces were collected from 20 calves (Trials 1, 2a, 3a) or 12 pens (Trials 2b, 3b) during each 5-d fecal collection period. Chromic oxide in MR and acid insoluble ash in CS were used as digestion markers before and after weaning, respectively. Apparent TTD of NDF, non-fiber carbohydrate (NFC), CP and fat were used to calculate ME in CS, which was subsequently adjusted for contributions of CP and fat from MR. Mixed model analyses for meta-analyses outlined by St. Pierre (2001; J. Dairy Sci. 84:741–755) were used to build prediction models. A total of 210 observations were used. Estimates of NFC and NDF TTD in CS were low early in life and increased with increasing CS nutrient intake (particularly NFC) in a logarithmic fashion. Because CS were composed mainly of NDF and NFC, calculated ME in CS was low, but increased with increasing NFC intake. Natural logarithm of daily NFC intake (lnNFCi, kg/d) accounted for 61, 12, 22, 57, and 87% of the variation in TTD of DM, CP, fat, NDF and NFC, respectively, in CS. Equation for ME (Mcal/kg DM) in CS was 0.3938 ± 0.0511 ´ lnNFCi + 2.9785 ± 0.0551 (r² = 0.70). Calculated ME in CS was < 75% of NRC estimated ME when calves consumed 0.2 kg/d of NFC, 98% of NRC estimated ME at 1 kg/d of NFC intake and 109% of NRC estimated ME at 2 kg/d of NFC. Ability of the calf to extract energy from CS is dynamic and depends on intake of NFC and NDF from CS. Feed management that delays CS intake may reduce TTD of NFC and NDF and calculated ME in CS.