Background. To promote early diagnosis and treatment of short stature, consensus meetings were held in the mid nineteen nineties in the Netherlands and the UK. This resulted in guidelines for referral. In this study we evaluate the referral pattern of short stature in primary health care using these guidelines, comparing it with cut-off values mentioned by the WHO. Methods. Three sets of referral rules were tested on the growth data of a random sample (n = 400) of all children born between 01-01-1985 and 31-12-1988, attending school doctors between 1998 and 2000 in Leiden and Alphen aan den Rijn (the Netherlands): the screening criteria mentioned in the Dutch Consensus Guideline (DCG), those of the UK Consensus Guideline (UKCG) and the cut-off values mentioned in the WHO Global Database on Child growth and Malnutrition. Results. Application of the DCG would lead to the referral of too many children (almost 80%). The largest part of the referrals is due to the deflection of height, followed by distance to target height and takes primarily place during the first 3 years. The deflection away from the parental height would also lead to too many referrals. In contrast, the UKCG only leads to 0.3% referrals and the WHO-criteria to approximately 10%. Conclusion. The current Dutch consensus guideline leads to too many referrals, mainly due to the deflection of length during the first 3 years of life. The UKCG leads to far less referrals, but may be relatively insensitive to detect clinically relevant growth disorders like Turner syndrome. New guidelines for growth monitoring are needed, which combine a low percentage of false positive results with a good sensitivity. © 2007 Grote et al; licensee BioMed Central Ltd.