Splitting or longitudinal shear failures in full-culm bamboo are a critical limit state when using the material for construction applications. Understanding and quantifying the behaviour must be through standardised test methods for international adoption into building practices to be successful. Current testing procedures only determine an effective flexural strength due to splitting failures. The present study investigates longitudinal shear failure in full-culm Phyllostachys pubescens (Moso) and Bambusa stenostachya (Tre Gai) bamboo flexural specimens through a number of standard and modified test methods. In particular, the paper proposes modifications to ‘standard’ culm flexural tests intended to better capture actual shear-dominated behaviour. Notched specimens were used to better establish relationships between mode I and mode II behaviours. The test results illustrate the significant degree of interaction between mode I and mode II stresses resulting from flexural tests and their associated failures, specifically the significant deterioration of mode II capacity in the presence of even a small amount of mode I stress. It is proposed that a uniform experimental approach using horizontally notched flexural specimens having different shear spans and notch locations would be appropriate to establish a relationship between mode I and II behaviour.