Objectives The aim of this study was to evaluate whether spinal canal depth (SCD), measured using ultrasound, could be estimated from simple body measurements in a sample of children.

Methods We measured SCD in a group of 225 children aged 0–18 years in the curved left lateral position using ultrasound. Statistical analysis was performed using Pearson's correlation coefficient at the 5% level of significance. We also performed linear regression analyses (analysis of variance) for mid-spinal canal depth (MSCD), including five potential predictors of age, gender, height, weight and body surface area, in each model.

Results The mean MSCD was 33.0 mm (18.1–56.4) across the whole cohort. The best linear correlation of MSCD (mm) was found with weight (W; kg), approximating the formula MSCD=0.4W+20 (R²=0.72). Body weight accounted for 85% of the variance in the data (adjusted R²=0.72). Our formula gives values outside of the actual measured SCD range in 23/225 (10.2%) of cases and estimates MSCD at 24 mm at 10 kg, 32 mm at 30 kg and 40 mm at 50 kg.

Conclusions We demonstrate a good correlation between weight and MSCD in a large group of children. Use of the simple formula MSCD (mm)=0.4 W+20 could improve the success rates of lumbar puncture in the paediatric population, but remains to be validated.