Adequate dietary habits are supposed to be one of the most important modifiable factors in osteoporosis prevention. However, the importance of specific nutrients is controversial. We examined relevant nutrients which are supposed to have an impact on bone parameters and compared their effect sizes with those of two known predictors of bone development: bone-related muscle mass and androgen levels. We analysed nutritional, hormonal and anthropometric data from 107 prepubertal children participating in the Dortmund Nutritional and Anthropometric Longitudinally Designed Study. Diaphyseal bone mineral content (BMC), cortical area (CA), periosteal circumference, strength strain index and muscle area of the non-dominant forearm were measured by peripheral quantitative computed tomography. Data on long-term nutrient intakes (e.g. protein, Ca and vitamin D) were derived from 3 d weighed dietary records. Twenty-four hour urinary excretion rates of androgen metabolites including the sex steroid androstenediol were measured using GC-MS. Of all considered nutrients, only protein showed a trend for an association with BMC (β = +0·11; P = 0·073) and CA (β = +0·11; P = 0·056) in stepwise linear regression models. None of the other considered dietary variables was associated with bone parameters. The size of the bone anabolic effect of protein was partly comparable with that of androstenediol. Even though boys gained more bone mass in comparison with girls, the protein effect did not differ between sexes. Bone-related muscle area and sex steroids have the strongest effects on prepubertal diaphyseal bone. However, dietary protein may have a similar bone anabolic influence compared with androstenediol. In children without explicit nutrient deficits, protein seems to be the most important dietary component for diaphyseal bone status.