In this project, we investigate whether the SNARC effect (Spatial-Numerical Association of Response Codes; Dehaene et al., 1993; faster responses to small/large numbers with the left/right, respectively) depends only on relative or also on absolute number magnitude. In two seminal studies (Dehaene et al., 1993; Fias et al., 1996), numbers 4 and 5 were associated with the right in the lower ranges where they were the largest numbers, but with the left in the higher ranges where they were the smallest numbers, speaking for the dependency on relative number magnitude. However, full flexibility was concluded from null results regarding the role of absolute number magnitude in frequentist analyses in these two poorly powered studies We ran two highly powered experiments to investigate whether Spatial-Numerical Associations are fully flexible or also (partly) fixed: - a **direct replication** of Dehaene et al. (1993) and Fias et al. (1996) using their number ranges from **0 to 5** and from **4 to 9** - a **conceptual replication** considering advances in SNARC research using number ranges from **1 to 5 excluding 3** and from **4 to 8 excluding 6** We found that not only relative, but also absolute number magnitude drives Spatial-Numerical Associations. Number 1 is strongly associated with the left, and the SNARC effect is stronger in lower than in higher number ranges. Spatial-Numerical Associations are thus partly flexible and partly fixed.