UNLABELLED: Testosterone plays a vital role in the development of boys. There is a lack of age-specific reference data for testosterone representing a gap in our abilities to differentiate physiological deviations from pathological conditions such as delayed or precocious puberty, anatomical aberrations such as hypospadias or epispadias, micro-penis or buried-penis and disorders of sex development. OBJECTIVE: To establish age-specific reference values for serum testosterone in boys aged 1-14 years and generate centile curves to predict testosterone levels by age. MATERIALS AND METHODS: Serum testosterone levels were measured in 1398 boys aged 1-14 years using chemiluminescent microparticle immunoassay. The Generalized Additive Models for Location, Scale, and Shape framework was used to generate centile curves (3rd, 5th, 10th, 25th, 50th, 75th, 90th, 95th and 97th percentiles) for testosterone as a function of age. RESULTS: The study revealed an age-dependent increase in serum testosterone (correlation coefficient:0.6
p ≤ 0.001). Serum testosterone remain low with minimal inter-individual variability until the onset of puberty, which is marked by a steep rise in levels. The inter-percentile range is widened significantly during puberty, indicating high variability in testosterone and co-existence of pre-pubertal boys with those already into puberty. CONCLUSIONS: The centile curves and reference values will bridge the literature-gap and serve as a critical tool in clinical practice. The findings also underscore the relevance of personalized assessment in view of individual variability in testosterone levels and the timings of the onset of puberty.