About 10% of hyperthyroid cats have a normal total T4 (TT4), requiring further testing to make the diagnosis. Thyroid stimulating hormone (TSH) is measured using the "canine" assay (TSH-CLIA, Immulite 2000 by Siemens) as the only assay currently available. However, this assay cannot differentiate between subnormal and low-normal TSH concentrations in cats due to poor specificity (70-85%). A novel feline-optimized TSH assay (TSH-BAW, Truforma by Zomedica) was recently developed. It allows differentiation between euthyroid and hyperthyroid cats. However, the effect of non-thyroidal illness (NTI) on TSH-BAW has not been evaluated. Our objectives included the comparison of serum TSH concentration using both the TSH-CLIA and TSH-BAW assays among hyperthyroid cats, cats with NTI, and healthy cats, and the evaluation of the sensitivity and specificity of the TSH-BAW for diagnosis of FHT. This prospective cross-sectional study was performed on 102 client-owned cats, including 37 hyperthyroid, 33 healthy, and 32 NTI cats. The following thyroid hormones were measured in all cats: TT4, TSH with both assays (Immulite 2000 and Truforma). Hyperthyroidism was confirmed by thyroid scintigraphy. Euthyroidism was confirmed by repeating TT4 measurement at least three months after enrollment (if available) to rule out subclinical hyperthyroidism. Cats with NTI were further divided based on the severity of their illness. Serum TSH was compared among groups using Kruskal-Wallis followed by Dunn's procedure, and compared among NTI severity scores using the Fisher's Exact test. Significance was set at P <0.05. The sensitivity and specificity of TSH-BAW for detecting hyperthyroidism are 78% (62-90%) and 97% (84-100%), respectively. The median TSH is significantly different between hyperthyroid cats and healthy and NTI cats with both assays (P<0.01). The TSH was not different between the latter euthyroid groups (P=0.87 and P=0.29). Eight (21.6%) hyperthyroid cats have a normal TSH-BAW but undetectable TSH-CLIA. Twelve (4 healthy, 8 NTI) euthyroid cats (18.5%) have an undetectable TSH-CLIA with only two (1 healthy, 1 NTI) (3%) having an undetectable TSH-BAW. The proportion of cats with a suppressed TSH is higher with severe illnesses with the TSH-CLIA only. In conclusion, the TSH-BAW has a high specificity, identifies normal TSH in healthy cats more often, and appears to not be affected by NTI. It can be a useful tool for the diagnosis of feline hyperthyroidism. However, a low-normal TSH cannot be used to rule out hyperthyroidism. / Master of Science / Hyperthyroidism leads to elevation of the thyroid hormone total T4 (TT4). About 10% of hyperthyroid cats have a normal TT4, requiring further testing to make the diagnosis. Another thyroid hormone, thyroid stimulating hormone (TSH), could be used like it is with people. In feline medicine, it is measured using the "canine" assay (TSH-CLIA, Immulite 2000 by Siemens) as the only assay currently available. However, this assay cannot differentiate between subnormal and low-normal TSH concentrations in cats due to poor specificity (70-85%). A novel feline-optimized TSH assay (TSH-BAW, Truforma by Zomedica) was recently developed. It allows differentiation between euthyroid and hyperthyroid cats. However, the effect of non-thyroidal illness (NTI) on TSH-BAW has not been evaluated. Our objectives included the comparison of serum TSH concentration using both the TSH-CLIA and TSH-BAW assays among hyperthyroid cats, cats with NTI, and healthy cats, and the evaluation of the sensitivity and specificity of the TSH-BAW for diagnosis of FHT. The study was performed on 102 client-owned cats, including 37 hyperthyroid, 33 healthy, and 32 NTI cats. The following thyroid hormones were measured in all cats: TT4, TSH with both assays (Immulite 2000 and Truforma). Cats with NTI were further divided based on the severity of their illness. Serum TSH was compared among groups using Kruskal-Wallis followed by Dunn's procedure, and compared among NTI severity scores using the Fisher's Exact test. Significance was set at P <0.05. The sensitivity and specificity of TSH-BAW are 78% and 97%, respectively. The median TSH is significantly different between hyperthyroid cats and healthy and NTI cats with both assays. The euthyroid cats (healthy and NTI cats) were not different. Eight (21.6%) hyperthyroid cats have a normal TSH-BAW (not normal in the face of hyperthyroidism) but undetectable TSH-CLIA. The proportion of euthyroid cats with a suppressed TSH (not normal in the face of euthyroidism) is higher with the TSH-CLIA compared to the TSH-BAW. Only with the TSH-CLIA, the proportion of NTI cats with a suppressed TSH is higher than healthy cats, and is higher with severe illnesses. In conclusion, the TSH-BAW has a high specificity, identifies normal TSH in healthy cats more often, and appears to not be affected by NTI. It can be a useful tool for the diagnosis of feline hyperthyroidism. However, a low-normal TSH cannot be used to rule out hyperthyroidism.