The factor structure, the concurrent validity, and test-retest reliability of the Czech translation of the Gifted Rating Scales-School Form [GRS-S; Pfeiffer, S. I., & Jarosewich, T. (2003). "GRS (gifted rating scales) - manual." Pearson] were evaluated. Ten alternative models were tested. Four models were found to exhibit acceptable fit and interpretability. The factor structure was comparable for both parent (n = 277) and teacher raters (n = 137). High correlations between the factors suggest that raters might be subject to a halo effect. Ratings made by teachers show a closer relationship with criteria (WJ IE II COG, CFT 20-R, and TIM[superscript 3-5]) than ratings made by parents. Test-retest reliability of teacher rating (with median 93 days) was quite high for all GRS-S subscales (r = 0.84-0.87).