Density is one of the fundamental properties of fibres which reinforce polymer matrix composites, and is used both to estimate composite weight and to evaluate fibre content for property predictions. For traditional composites, reinforced by glass or carbon fibres, unique density values are well known for a particular fibre grade and provide reliable fibre content estimations and composite property predictions. However, this is not the case for natural fibres. This paper first reviews published density values for flax (Linum usitatissimum L.) fibres and describes the limitations of techniques used to measure fibre density. Significant variations in published densities are found, which can be related to the measurement method. New data quantify the influence of measurement technique, fibre extraction route, moisture content, and reinforcement geometry. Values obtained by buoyancy are around 10% lower than those obtained by pycnometry. Finally the consequences of measured density variations on property estimations for long fibre composites are discussed; volume fraction values calculated from buoyancy provide more accurate tensile modulus values compared to experimental values than those from gas pycnometry; the former are recommended for predictive use.