Objectives: To determine cutoff values in small (SB) and medium/large (MLB) breed dogs with and without medial patellar luxation (MPL) for identifying abnormal femoral trochlea morphology. Study design: Original research.Animals: A total of 80 computed tomographic (CT) scans from client-owned dogs Methods: Four groups of 20 dogs were created: (1) control SB, (2) control MLB, (3) MPL-SB, and (4) MPL-MLB. Two authors measured the femoral trochlear groove angle (FTGA), femoral trochlear angle (FTA), and femoral trochlear ridge inclination angle (FTRIA) in two points with CT. ANOVA and ROC-analysis were tested to the control and MPL groups to assess sensitivity, specificity, and cutoff values. Statistical significance was set to p < .05. Intraclass correlation coefficients evaluated the inter-rater agreement. Results: FTGA (± SD) in control SB (128.8 ± 4.7 ) and control MLB (119.2 ± 5.6 ), was smaller (p < .0001) than in MPL-SB (139.4 ± 4.4 ) and MPL-MLB (133.7 ± 5.1 ). FTA and FTRIA were decreased (p = .12, p = .23) in MPL-SB (2.1 ± 6.8; À0.3 ± 3.3 ) and MPL-MLB (3.8 ± 5.6 ; 1.7 ± 4.5 ) compared to control SB (0.2 ±4.1; À0.1 ± 2.6 ) and control MLB (5.3 ± 2.8 ; 3.1 ± 1.3 ).Cutoff values for FTGA, FTA, and FTRIA were > 134 , < À5.9 , < À2 (SB), and > 128.3 , < À0.4 , < À0.4 (MLB). Sensitivity, specificity, and inter-rater agreement were superior for FTGA than FTA and FTRIA.Conclusions: Dogs without MPL had a deeper femoral trochlear groove than MPL dogs. SB had a shallower groove than MLB. The measurement of FTA and FTRIA was not reliable.Clinical relevance: A FTGA <134 (SB) and < 128 (MLB) may be considered as a cutoff for trochleoplasty decision-making.