AbstractA study was carried out to determine the effect of replacing the cake portion of concentrate mixture with 5 parts polyethylene glycol (PEG)‐6000 in a complete feed block (CFB) containing Prosopis cineraria leaves on the performance, rumen fermentation pattern and rumen enzyme profile of kids under an intensive system of rearing. Eighteen weaners of Sirohi goat of similar body weight (16.0 ± 0.5 kg) and age (90 ± 5 days) were divided into three equal groups. They were housed in individual cages in a side open asbestos roof shed with mud floor. All the kids received CFBs containing 50 parts P cineraria leaves and 50 parts concentrate mixture. CFB offered to the first group had high protein (183.8 g kg−1; HP), that offered to the second group had low protein (131.3 g kg−1; LP) and that offered to the third group had low protein (124.7 g kg−1) but contained PEG‐6000 (LP‐PEG). The concentrate mixtures in LP and LP‐PEG were without groundnut cake, whereas in LP‐PEG, groundnut cake was replaced by barley and 5 parts PEG‐6000 were incorporated. CFBs were similar in their nutritive value except for crude protein (CP). Prosopis leaves utilized in the CFB contained (g kg−1 dry matter) 129.1 CP, 535.5 neutral detergent fibre, 395.8 acid detergent fibre and 222.8 acid detergent lignin. There were significant differences in dry matter intake (g day−1) between HP (1102), LP (1108) and LP‐PEG (1194); the trend in Metabolizable energy intake was similar. During the growth trial, LP kids consumed maximum amount of feed (76.91 kg) followed by HP (75.73 kg) and LP‐PEG (73.12 kg). However, maximum feed efficiency (feed consumed kg−1 live weight gain) was recorded in LP‐PEG kids (9.59) followed by HP (10.64) and LP (11.60). These differences were statistically significant (p < 0.05). Although there was no significant difference in the digestibility of dry matter among the groups, there was significant difference in the digestibility of CP, neutral detergent fibre and acid detergent fibre. The digestibility of CP was 0.591, 0.484 and 0.645, that of neutral detergent fibre was 0.397, 0.308 and 0.499 and that of acid detergent fibre was 0.168, 0.154 and 0.282 in HP, LP and LP‐PEG, respectively. Rumen metabolites studied 6 h after feeding revealed that there were significant (p < 0.05) differences in the concentrations of ammonia N, tri‐chloro acetic acid precipitable N and total volatile fatty acids among the three groups, but not pH. The rumen enzyme concentrations showed significant (p < 0.05) differences for α‐amylase, whereas the differences were non‐significant for CMcase and protease. After 90 days of feeding trial, the maximum weight gain was recorded in LP‐PEG (7.62 kg) followed by HP (7.23 kg) and LP (6.53 kg). It can be concluded that, when kids are reared under an intensive system on complete feed blocks containing Prosopis leaves, high protein concentrate can be replaced with a low protein concentrate containing 5 parts PEG, which would not only alleviate the negative effects of tannin but would also spare expensive groundnut protein. Copyright © 2005 Society of Chemical Industry