The objective of this study was to determine evaluator effect on genetic predictions of docility score (DS), temperament score (TS), and qualitative behavior attributes (QBA; 6 positive and 6 negative QBA). Weaning age calves (n = 1,542) were evaluated using DS, TS, and QBA scores over 4-year period. Fixed effects included evaluator (n = 11 total), primary breed (n = 2), sex (n = 2), and random effect of calf using repeated measure design (across-evaluator model only). Variance components, heritability, and breeding values (EBV) were estimated using pedigree in ASReml 4.2. Evaluator effect on EBV was based on 1) Spearman rank correlation coefficients (rs) and 2) 3-quartiles change in rankings of calves among evaluators per trait. Results showed evaluator scoring was different in DS, TS, and 12 QBA (P > 0.002). Correlations (rs) on EBV across and within evaluators ranged from 0.52 to 0.84 (DS) and 0.53 to 0.87 (TS) (P > 0.0001). For QBA, all rs on EBV were significant (P > 0.05) and ranged from -0.26 to 0.90 (positive) and 0.11 to 0.82 (negative), except on positively occupied and distressed QBAs (P = 0.46 to 0.65). When comparing evaluator specific EBV to across evaluator model, DS and TS had 1.10% and 1.56% 3-quartiles change, whereas comparison of evaluator to evaluator EBV showed 4.54% and 3.76%. For positive and negative QBA attributes, 3-quartiles change ranged from 0% to 35.13% and 0.52% to 3.76% (evaluator to across-evaluator model); and 1.30% to 21.09% and 3.31% to 13.81% (within evaluators). This illustrates that DS and TS had less variability in EBV prediction compared to QBA. Within QBA, negative QBA had less variability compared to positive QBA. In conclusion, evaluators scored differently using these methods. However, in predicting genetic merit, evaluator has negligible effect for scoring systems already implemented by breed associations.