Body depth less than head length, depth contained 3.0 to 3.4 times in standard length (for fish 10 to 46 cm standard length). Head length contained 2.5 to 2.7 times in standard length; interorbital area convex; preopercle angular, with 2 or 3 greatly enlarged serrae at the angle; in adults larger than 40 cm standard length, the preopercle angle is produced into a rounded lobe, with an indentation immediately above the lobe; middle and lower opercular spines flat but distinct, upper spine not apparent; upper edge of operculum straight or slightly convex; nostrils subequal in specimens less than 30 cm standard length; rear nostril diameter about twice that of anterior nostril in fish of 40 to 50 cm standard length; maxilla usually reaching a vertical at rear edge of eye; ventral edge of maxilla with a low step; no scales on maxilla; midlateral part of lower jaw with 2 rows of teeth. Gill rakers 8 to 10 on upper limb, 16 to 18 on lower limb, total 24 to 27. Dorsal fin with XI spines and 15 to 17 rays, the third or fourth spine longest, subequal to longest dorsal-fin ray, the interspinous membranes distinctly incised; anal fin with III spines and 8 soft rays; pectoral-fin rays 18 or 19; pectoral fins usually longer than pelvic fins, pectoral-fin length contained 1.6 to 2.1 times in head length; caudal fin truncate or slightly convex in juveniles, becoming concave or lunate in adults larger than 40 cm standard length. Lateral-body scales ctenoid; adults with auxiliary scales; lateral-line scales 70 to 73; lateral-scale series 113 to 130. Pyloric caeca 17.
Colour: Head and body brownish, the fins darker. Juveniles less than 15 cm standard length with 3 to 5 narrow dark stripes (possibly blue in life) paralleling lateral line on dorsal part of body: 2 stripes above and 1 to 3 stripes below lateral line. Two dark lines on head: one from lower edge of eye to ventral rear edge of interopercle, the second from dark maxillary streak to lower edge of preopercle. Adults brown or greyish brown, often with a large, distinct golden yellow blotch (vaguely defined at periphery) on body below spinous dorsal fin. Two specimens from Angola, (Museu Bocage nos. MB 2087 and 2091, 46 and 42 cm standard length; reported by Franca, 1957) are distinctly bicoloured, the body dark brown dorsally and abruptly paler ventrally, the two parts separated by a wavy boundary. Both fish are males, with flaccid testes containing a large empty lumen; if the condition of the testes is indicative of recent spawning, the bicoloured pattern may be the spawning coloration of this species.

E. costae occurs in the eastern Atlantic anti Mediterranean. Heemstra and Randall (1993) have examined specimens from Greece (Corfu Island), the Cape Verde Islands, and Angola. Reliable literature records document its occurrence on the Mediterranean coasts of Italy, France, Spain, Egypt, Tunisia, also along the south coast of Portugal, and along the west coast of Africa from Morocco to southern Angola. Alberto Brito informed Heemstra and Randall (1993) that E. costae is known in the Canary Islands from only 10 records, and these are mostly based on juvenile specimens. Records of “Epinephelus alexandrinus” from Madeira (see Remarks, below) are apparently based on misidentifications of Mycteroperca fusca.

According to Poll (1954) E. costae (identified as E. zaslavskii) is found on a variety of bottom types (sand, mud, or rock) in depths of 20 to 80 m. Maurin (1968) reported Epinephelus alexandrinus (probably E. costae) as rather common on clumps of coral off Cape Blanc on the north coast of Mauritania. Age and growth of Egyptian specimens (identified as “Epinephelus alexandrinus”) were studied by Wadie et al. (1981). Bouain (1986) reported a maximum age of 11 years for “Epinephelus alexandrinus” of 50 cm standard length from Tunisia, and a growth coefficient (K) = 0.042 derived from the Von Bertalanffy growth equation for his population. Bouain and Siau (1983) estimated total potential fecundity for an “Epinephelus alexandrinus” of 46.5 cm standard length to be 879 038 eggs. Larvae described as “Epinephelus alexandrinus” by Bertolini (1933) and Sparta (1935) are actually Mycteroperca rubra.

Maximum total length probably about 80 cm. According to Tortonese (1986), this species (identified as “Epinephelus alexandrinus”) attains 140 cm standard length.

Following Boulenger’s (1895) authoritative work on serranid fishes, this species has generally been referred to as Epinephelus alexandrinus (Valenciennes, 1828). Unfortunately, Valenciennes' holotype of Serranus alexandrinus (MNHN 7325) is a specimen of the well-known Epinephelus fasciatus of the Red Sea and Indo-Pacific region. This holotype differs significantly from the species here recognized as E. costae in having fewer scales (lateral-line 54, versus 70 to 73; lateral-scale series about 100, versus 113 to 130), deeper body (depth 2.8 times in standard length, versus 3.0 to 3.4 times in standard length), 3 or 4 rows of teeth at midside of lower jaw (versus 2 rows), fewer gill rakers (7 on upper limb and 1.5 on lower limb, 22 in total, versus 8 to 10 on upper limb, 16 to 18 on lower limb, 24-27 in total); and a rounded caudal fin (the shape of the caudal fin of the holotype cannot now be determined, as it is damaged; but in his description of Serranus goreensis, Valenciennes (1830) mentioned that the caudal fin of E. alexandrinus is rounded). In his original description, Valenciennes (1828) wrote “Sa couleur parait avoir été brune, sans taches ni marbrures, sur tout le corps et sur les nageoires.” E. fasciatus has distinctive black triangles at the margin of the interspinous dorsal-fin membranes, and these are clearly seen on the holotype of S. alexandrinus if the fin is erected. This feature was overlooked by Valenciennes and subsequent workers. The holotype also still shows the dark pigment on the edge of the orbit that is typical of E. fasciatus. Although Valenciennes gave the place of origin of his holotype as “rapportée de I’Egypte par M. Geoffroy,” his choice of name for his new species (Serranus alexandrinus) implied that it was a Mediterranean species; and this accounts for the misapplication of this species name by subsequent authors. Bauchot et al. (1960) discussed the synonymy of "Epinephelus alexandrinus" in which they included Epinephelus zaslavskii Poll, 1949, but they did not give any data on the holotype of Serranus alexandrinus. E. costae is similar to E. goreensis in meristic and most morphometric features, but these two species can be distinguished by their colour patterns. Juveniles of E. costae have 3 to 5 narrow dark stripes paralleling the lateral line (no dark stripes in E. goreensis), and live (or freshly dead) adults of E. costae often have a large golden yellow blotch on the body below the spinous dorsal fin. E. costae has never been reported with the subvertical dark bars or “saddles” on the dorsal part of the body that are characteristic for E. goreensis. Séret (1981) published an excellent figure of E. costae (identified as E. goreensis). Reports of "Epinephelus alexandrinus” from Madeira and the Azores (Saldanha, 1979; Waschkewitz and Wirtz, 1990) are apparently misidentifications (based on underwater visual identifications or photographs of live fish) of Mycteroperca fusca, which is superficially similar to E. costae (both species are relatively elongate, somewhat compressed groupers with concave or lunate caudal fins in adults and a protruding lower jaw). The Spanish common name “falso abadejo” for E. costae alludes to its similarity to M. fusca, the true “abadejo.” M. fusca is common at Madeira, and is well known to the fishermen as “abadejo.”According to G.E. Maul, ichthyologist at the Funchal Municipal Museum for the past 50 years, M. fusca and the mero (E. marginatus) are the only two species of groupers that occur in Madeiran waters.Examination of Madeiran specimens in the Funchal Museum and at the British Museum (Natural History) also confirms Mr. Maul’s statement.