Abstract Background Metatarsus adductus is a common congenital foot deformity. Variable prevalence values were reported using different techniques in different populations. Numerous radiological measurements have been proposed to assess this deformity with a paucity of studies reporting the reliability of these methods. The metatarsus adductus angle was shown to correlate with the severity of hallux abductovalgus in normal feet and preselected populations of juvenile hallux valgus. Materials and methods Weight bearing dorsoplantar radiographs of 150 feet were examined for 5 angles commonly used in assessing metatarsus adductus: angle between the second metatarsus and the longitudinal axis of the lesser tarsus (using the 4th or 5th metatarso-cuboid joint as a reference), Engel's angle and modified Engle's angle. The prevalence of metatarsus adductus was assessed according to published criteria for different techniques. Inter and intra-observer reliabilities of these angles were evaluated on 50 X-rays. Linear regression tests were used to assess the correlation between hallux valgus and different angles used in assessing metatarsus adductus. Results Intraclass correlation coefficients were high for intra- as well as inter-observer reliability for the 5 angles tested. Prevalence of metatarsus adductus ranged (45–70%) depending on the angle used in the same population. Only the metatarsus adductus angle using the 4th metatarso-cuboid joint as a reference demonstrated significant correlation between metatarsus adductus and hallux abductovalgus angles. Conclusion Five techniques commonly used in assessing metatarsus adductus demonstrated high inter and intra-observer reliability values. Prevalence of metatarsus adductus and the correlation between the severity of this deformity and hallux valgus angle is sensitive to the assessment method.