One thousand and six hundred 1-day-old sexed broiler chickens (Vedette; 50% male, 50% female) were randomly allocated to eight treatment groups in order to test the effect of the ROXAZYME G enzyme complex (beta-glucanase, xylanase, amylase, pectinase) in pelleted and extruded high-energy diets on broiler performance. Two diets were tested. The first was a commercial diet based on wheat and com, and in the second 20% wheat was replaced by barley. Both diets were tested in pelleted and extruded form with or without enzyme which was added before processing at the level of 0.15 g kg-1. The eight treatments tested were as follows: Treatments A and B, pelleted wheat-com diets without and with enzyme respectively; Treatments C and D, extruded wheat-com diets without and with enzyme; Treatments E and F, pelleted barley-wheat-com diets without and with enzyme; Treatments G and H, extruded barley-wheat-com diets without and with enzyme. Energy, fat, N, and fibre utilisation were determined with four diets containing barley in the 5th week of growth. The enzyme complex significantly improved feed conversion ratio (FCR) (1.81 vs. 1.77; P<0.001) until the end of the growing period (39 days), and liveweight at Day 21 (0.66 kg vs. 0.68 kg; P<0.001). Highly significant differences were found for liveweight and feed conversion at Day 39 between diets (20% wheat vs. 20% barley: liveweight, 1.81 kg vs. 1.75 kg; FCR, 1.76 vs. 1.81), and also between processing (pelleting vs. extruding: liveweight, 1.84 kg vs. 1.73 kg; FCR, 1.76 vs. 1.82). Birds fed on extruded feeds grew slower and utilised feed less efficiently than birds fed on pelleted feeds. The extrusion process increased the soluble fibre content in the feed, and subsequently the feed extract viscosity was also increased. The interaction between processing (pelleting vs. extrusion) and grain (wheat vs. barley) for liveweight at Days 21 and 39 was significant, indicating a greater negative impact of extrusion on broiler feed containing barley. In the metabolism studies a positive trend caused by enzyme supplementation was seen on energy metabolisability, fat, N, and fibre utilisation, but the differences were not significant.