Objective: To assess the variables useful to predict caesarean delivery (CD) and instrumental assistance, through the analysis of a large number of foetal-pelvic variables, using discriminant analysis. Materials and Methods: One hundred and fourteen pregnant women were included in this single-centre prospective study. For each mother-foetus pair, 43 pelvic and 18 foetal variables were measured. Partial least squares-discriminant analysis was performed to identify foetal-pelvic variables that could statistically separate the 3 delivery modality groups: spontaneous vaginal delivery (SVD), CD, and instrument-assisted delivery (IAD). Results: For the SVD versus CD model, voluminous foetuses and women with a narrow pelvic inlet had a greater risk for requiring CD. The most efficient variables for discrimination were the transverse diameter and foetal weight. The antero-posterior inlet and obstetric conjugate were considered in this model, with the former being a useful variable but not the latter. For the SVD versus IAD model, the most important variables were the foetal variables, particularly the bi-parietal diameter. Women with a reduced antero-posterior outlet diameter and a narrow pubic arch were more at risk of requiring an IAD. Conclusion: The antero-posterior inlet was an efficient variable unlike the obstetric conjugate. The obstetric conjugate diameter should no longer be considered a useful variable in estimating the arrest of labour. Antero-posterior inlet diameter was a sagittal variable that should be taken into account. The comparison of sub-pubic angle and bi-parietal and antero-posterior outlet diameters was useful in identifying a risk of requiring instrumental assistance.