Behavioural disorders in dogs are common and have severe welfare consequences for dogs. This study aimed to assess the factors that are significant and predictive of behaviour problems in dogs using the animal welfare assessment grid (AWAG) to further understand what factors influence their welfare. 177 AWAG assessments were undertaken across 129 dogs that clinicians deemed to have a behavioural disorder. Wilcoxon rank-sum tests were used to assess the difference in scores between dogs with behaviour disorders and a cohort of healthy dogs (n = 117). This analysis showed that all physical factors besides body condition, all procedural factors besides procedure pain, and all psychological, and environmental factors were significantly different between healthy dogs and dogs with behaviour disorders. Spearman rank correlation coefficient (RS) revealed several significant strong positive correlations including the procedural impact on the dog’s daily routine with aggression towards unfamiliar people and procedure pain, as well as other correlations between the dog’s behaviour during assessment with the frequency at which they encounter fears and anxieties, clinical assessment and procedure pain, and reaction to stressors and social interactions. These findings highlight the interdependent nature of the various influences of welfare. Logistic regression analysis identified that aggression towards the caregiver, fears and anxieties frequency, and choice, control, and predictability were all significant predictors of behaviour disorders. The findings have important implications for veterinary, behaviour, and animal welfare professionals as any changes across these factors may indicate poor welfare linked to emotional disorders in dogs.