To evaluate the raw material digestibility, the physical characteristics of diet pellets and faeces, and the postprandial nitrogen excretion in greater amberjack (Seriola dumerili) reared in recirculating aquaculture systems (RAS), 10 diets were obtained by mixing a reference diet (CP: 51.8% DM; EE: 13.6% DM) with one of ten raw materials (7:3 ratio), i.e. fishmeal (FM), porcine haemoglobin (PH), fermented soybean (FS), soybean protein concentrate (SPC), hydrolysed wheat protein (HWP), potato protein (PP), beet pulp (BP), beer yeast without fat (BY), beer yeast with fat (BYF), and Lithothamnium seaweed (LS). Twenty fish (268.9 ± 94.3 g) were randomly distributed into 5 RAS tanks (4 fish per tank) and fed for 140 days the experimental diets. A Latin square experimental design was used based on 5 diets × 5 tanks for two rounds (10 ingredients, 2 weeks per diet per tank). The highest ADCs of dry matter and protein were recorded for PH (83.0% and 88.4%, respectively), FM (81.8% and 84.3%), SPC (77.4% and 82.5%), and BYF (73.0% and 80.9%) while the lowest ones for BY (64.4% and 70.8%), HWP (59.8% and 69.9%), PP (48.3% and 69.8%), and BP (53.7% and 55.8%) (p < 0.05). The ADCs of EAA were the highest for PH, FM and SPC compared to the other raw materials; ADCs of lipids was above 90% for all raw materials (p > 0.05). Regarding pellets properties, FM was associated with the highest value for oil leakage whereas BP to the lowest (0.18% vs. 0.11%; p < 0.05); FS, HWP, BYF, and PH produced less turbidity in water after 15 min than BY (3.68 mg L−1 on average vs. 6.01 mg L−1; p < 0.05). Regarding faeces size, fish fed FS, BYF, and FM produced the lowest volume of fresh faeces of fine-size class than those fed the other raw materials (62.1% vs. 69.2% on average; p < 0.05); the highest rate of dry faeces retained at a sieve mesh interval of 1.0 < mm < 1.2, was produced by fish fed BYF and the lowest rate by those fed BY (1.16% vs. 0.34%; p < 0.05). A maximum of postprandial nitrogen excretion (mg N-NH4+ × kgfish−1 × h) was reached between 2 and 4 h after feed ingestion without differences among diets (p > 0.05). Overall, the findings of this investigation will contribute to the sustainable formulation of RAS aquafeed optimized and tailored for this promising Mediterranean aquaculture species.