Background Simultaneous measurement of 25(OH)D and 24,25(OH)2D is a new tool for predicting vitamin D deficiency and allows evaluating CYP24A1 lack of function. Interpretation of 24,25(OH)2D should be performed according to 25(OH)D levels and a ratio, called the vitamin D metabolite ratio (VMR) has been proposed for such a purpose. Unfortunately, the VMR can be expressed in different ways and cannot be used if 24,25(OH)2D concentrations are undetectable. Here, we propose evaluating the enzyme activity taking into consideration the probability that a normal population presents undetectable 24,25(OH)2D concentrations according to 25(OH)D levels. We thus retrospectively measured 25(OH)D and 24,25(OH)2D in a population of 1200 young subjects to evaluate the 25(OH)D threshold above which the enzyme was induced. Methods Serum samples from 1200 infants, children, adolescent and young adults were used to simultaneously quantify 25(OH)D and 24,25(OH)2D by LCMS/MS. Results Median (interquartile range [IQR]) levels were 20.6 (14.4–27.2) ng/mL for 25(OH)D. 172 subjects (14.3%) presented 24,25(OH)2D values below the LOQ. When 25(OH)D values were <11 ng/mL, 63.1% of subjects presented undetectable 24,25(OH)2D concentrations. Percentage decreased with increasing 25(OH)D values to become 19.7% for 25(OH)D comprised between 12 and 15 ng/mL, 5.1% for 25(OH)D between 16 and 20 and 0.7% for 25(OH)D >21 ng/mL. Conclusions We suggest using a statistical approach to evaluate CYP24A1 function according to 25(OH)D concentrations. Our results also show that vitamin D deficiency, as defined biochemically, could be around 20 ng/mL in infants, children, adolescent and young adults and that vitamin D deficiency could be evaluated on a more individual basis.