INTRODUCTION:Investigating the correlations between external pelvimetry variables and actual birth canal dimensions has important implications for clinical practice, especially for predicting dystocia. This validation study tests external and internal pelvimetry correlations.METHODS:This single-centered retrospective study included 142 women with records of computed tomography pelvimetry. The correlations between four internal variables (obstetric conjugate, transverse diameters of inlet, midplane, and outlet) and five external variables (intertrochanteric, inter anterosuperior iliac spines, external conjugate, transverse diagonal, and prepubic) were first studied individually. Then, the correlations between each internal variable and the seven external variables were jointly evaluated after adjusting for body mass index.RESULTS:As expected, the external conjugate correlated with the obstetric conjugate (r= .65;p< .01) but only weakly with the transverse outlet (r= .21;p< .05). In the simple correlation analysis, the intertrochanteric measure is also correlated with the transverse inlet (TRi), midplane (TRm), and outlet (TRo;r =.542–.672). The transverse diagonal measure marginally correlated with the TRi (r= .29;p< .01). In the multiple regression models, the intertrochanteric measure is associated with the TRi, TRm, and TRo, while the transverse diagonal is associated with the TRi (B= .27;p= .01). The prepubic diameter was not a predictor of birth canal dimensions in either simple or multiple regression models.CONCLUSION:Our study confirms that the reduction of the birth canal in some dimensions (i.e., obstetric conjugate and transverse diameters) is predictable with external, accessible pelvic diameters (e.g., external conjugate, transverse diagonal, and intertrochanteric).