Background Junior soccer players aged 16 to 19 years old competing on high level already present with postural deviations which are associated with an increased injury risk due to mechanical overloading of the neuro-musculoskeletal system. The application of valid postural assessments methods play an important role in terms of providing correct therapeutic consequences.

Objective Previous research show several methodological weaknesses and inconsistent results assessing posture through photographs. Thus, the purpose of the study was to determine the more conservative inter-rater reliability assessing the static posture involving the complete kinetic chain. It was hypothesised that photographic postural assessment is a reliable method.

Design Experimental, prospective cohort study.

Setting Qualitative assessment in clinical practice.

Patients (or Participants) 87 male elite junior soccer players out of three different age groups: Under 16, 17, and 19.

Interventions (or Assessment of Risk Factors) Athletes were photographed from four perspectives using professional photo equipment. An assessment protocol was completed for each player by three different highly qualified raters.

Main Outcome Measurements 37 variables describing postural characteristics, each variable graded by minimum of three, maximum of seven values.

Results Inter-rater reliability was fair only for the type of posture on side view with Kendal's W=0.6. For all remaining 36 aspects of postural characteristics Kendal's W was zero.

Conclusions Qualitative posture assessment is not valid and therefore not appropriate for diagnostic objectives, in particular for large scale based studies, e.g. for a single soccer team. For the purpose of meaningful postural evaluation qualitative assessments must be replaced by quantitative methods. Three-dimensional motion capture could be an appropriate alternative, avoiding radiation exposure by X-ray photography, since excellent agreement has been shown previously for different clinically-relevant kinematic parameters between these two methods.