The analgesic efficacy of meloxicam and ketoprofen against equine visceral pain is unclear. The aim of this study was to compare the analgesic efficacy of meloxicam (M) and ketoprofen (K) to flunixin meglumine (F) following inguinal castration. Horses undergoing inguinal castration under general anesthesia were randomly assigned F (1.1 mg/kg), M (0.6 mg/kg) or K (2.2 mg/kg) intravenously two hours pre-operatively and 24 h later. A pain score (out of 31) was recorded blindly by a senior clinician and veterinary student before NSAIDs administration (T0), and after the first (T1) and second (T2) administrations, using a modified post-abdominal surgery pain assessment scale (PASPAS). Pain was classified as mild (score ≤ 7), moderate (score = 8-14) or severe (score > 14). Thirty horses (12 F, 10 M, 8 K) aged 6.2 ± 4.9 years, mostly warmbloods, were included. Horse welfare was not compromised regardless of the drug assigned. There was no statistically significant effect of NSAIDs on pain score. Mean pain scores were significantly higher at T1 than T0 for each NSAID (F: 5.08 ± 2.50 vs. 1.58 ± 1.38 (p < 0.001); M: 4.60 ± 2.32 vs. 1.10 ± 1.20 (p < 0.001); K: 5.25 ± 1.39 vs. 1.50 ± 1.51 (p < 0.0001)) and lower at T2 than T1 for F (2.92 ± 2.423 vs. 5.08 ± 2.50 (p < 0.001)) and M (2.90 ± 1.37 vs. 4.60 ± 2.32 (p < 0.0325)). At T1, senior pain scores were significantly different than for junior (5.56 ± 0.54 vs. 3.22 ± 0.62, p = 0.005). This study indicates that meloxicam and ketoprofen provide a similar level of analgesia to flunixin meglumine for the management of mild visceral pain in horses. PASPAS is not reliable for junior evaluators.