isolates collected from diseased fish that were cultured along the south China coast. The antibiotic resistance profiles of 84 strains were tested with 12 common antibiotics using Kirby–Bauer disc diffusion methodology. The antimicrobial susceptibility data were used to identify correlations between antibiogram and subgroup types using cluster analysis software (SPSS 19.0, Statistical Product and Service Solutions). Eighty-four isolates formed 26 antibiogram types with an antibiogram abundance of 31.0%. The number of multiple antibiotic resistant types (i.e., resistant to more than three antibiotics) in the strains was 13 and comprised 50.0% of the total antibiogram types, causing the multi-antibiogram abundance value to increase to 59.1%. Isolates originating from Hainan, Guangdong, Guangxi, and Fujian Provinces possessed 18, 15, 3, and 1 of the antibiogram types, respectively, and all shared the J antibiogram type (FUR/AMO). Representative antibiogram types were J and M from 2007, A and B were added in 2010, 24 A to X types (A to X except for Y, Z) were added in 2011, while Y and Z appeared in 2012. The kinds of suppressive antimicrobial agents was 2, 5, 7, and 8, thereby mirroring the sampling years, 2007, 2010, 2011 and 2012, respectively. Strains isolated from Plectropomus leopardus, ♀ contained 6–10 antibiogram types, while the remaining fish sampled had only 1–3 types. The strains clustered initially into six subgroups (i–vi) and further formed groups I and II. Each subgroup contained characteristic antibiogram types comprising N, P, T, U, Y; R, S; K, P; F, H, O, X, V, W; E, G, Q; and C, L, Z. Common antibiogram types were not found in the Group I and II strains. Our research revealed that the isolates displayed polymorphic antibiogram types of moderately high abundance and could be typed for resistance effectively by cluster analysis. The resistance patterns of the isolates gathered from different sources were diverse.