Male Siamese fighting fish were held for 7 to 10 weeks either in visually-social conditions (interact aggressively with a mirror image for 2 min, three times a week) or as visually-isolated fish (no aggression, saw only the back of a mirror). Phase I of this experiment investigated whether visually-isolated fish became hyper-aggressive, compared to visually-social fish. Aggression levels were tested with a series of models and mirrors. Social isolation did not affect duration or frequency of specific agonistic behaviors, nor of pooled behavioral categories (ANOVA, P > 0.05) within aggression tests. Phase II was to determine if aggression levels re-adjusted in visually-isolated fish experiencing live encounters among groups of 4 fish in naturalistic environments. Data was not significant unlike in previous isolation experiments (ANOVA, P > 0.05). However, several behavioral trends were observed that were consistent with previous isolation studies.