The SNARC effect, i.e., the association of small/large numbers to the left/right side of space respectively, is widely studied in adults and generally observed at the group level in about 67% of the participants. However, Cipora et al.,(2019) showed that when using a bootstrapping approach to analyze the consistency of the SNARC effect at the individual level, its prevalence decreased to about 40%. To date studies on the SNARC effect in children are scarce and, so far, none explored it at the individual level. Therefore, our aim was (1) to test whether preschoolers show a SNARC effect at the group level and (2) to determine the prevalence of children with a consistent SNARC effect. Based on the methodology suggested by Cipora et al., (2019), we tested 136 preschool children (Mean Age: 6.3 years old) in the final year of kindergarten with a magnitude judgment task. Descriptively and at the group level, 67% of the children revealed a regular SNARC effect (p<.001). Nevertheless, bootstrapping analysis, with an 80% confidence interval, showed that the prevalence of children with a consistent regular SNARC decreased to 37% and that even 19% had an inverted consistent SNARC. The prevalence of regular SNARC is thus equivalent to the one observed in adults when tested with a parity judgment task. Our results suggest that access to a spatially oriented numerical representation emerges at an early developmental stage, with, however, only a minority consistently showing the SNARC effect.