In many species, low birth weight is identified as a major determinant for neonatal survival. The objectives of the present study were (i) to assess, in a large feline purebred population, the impact of birth weight on 0–2 months mortality in kittens, and (ii) if such mortality occurs, to define cut-off values for birth weight to identify at-risk kittens. Data from 5596 kittens from 15 breeds and provided by 194 French breeders were analysed. A logistic mixed model was used to identify low birth weight, being a male, and being born in a large litter as significant risk factors for kitten mortality during the first two months after birth. Classification and regression tree analysis was used to define the thresholds, first at the species level and, when possible, at the breed level. Two thresholds were defined to group kittens into three categories: low, moderate, or high risk of 0–2 months mortality (normal, low, and very low birth weight, respectively). In our population, 19.7% of the kittens were classified as low birth weight and 1.9% as very low birth weight. Critical thresholds may differ between breeds with similar birth weight distributions and equivalent mortality rates (e.g., Russian Blue/Nebelung vs. Egyptian Mau). These critical birth weight thresholds, established in 15 breeds, could be used to identify kittens requiring more intensive nursing to improve survival.