Subsea pipelines are crucial to the operations of the oil and gas industry, spanning many thousands of kilometres and extending as far as nearly 3000 metres in depth. The use of fibre-reinforced composite materials in deep-water wells has the potential to reduce operational costs due to their corrosion resistance and specific strength and stiffness relative to metals and their alloys. The reliability of composites is a limitation due to the complexity of damage occurrence caused by their inherent heterogeneous, layered nature. Structural health monitoring techniques have been the subject of many research studies, but there is limited literature available that focuses on these methods applied to tubular composites. This study aimed to make use of multiple different active and passive structural health monitoring techniques for damage detection and monitoring in tubular fibre-reinforced composites throughout the full life cycle. In particular, three methods were utilised: (i) in situ strain monitoring using distributed optical fibre sensors, (ii) active guided waves excitation in a pitch-catch configuration, and (iii) passive acoustic emission monitoring, using piezoelectric wafer active sensors. A considerable amount of literature is available on the application of these methods in metallic and composite structures, but the majority of research focuses on the use of a single method. In the present research, the methods are combined and applied to composite plates, composite tubes, and hybrid metal/composite tubular joints, for monitoring of the manufacturing processes and damage occurring during mechanical loading experiments. A novel unsupervised clustering approach is presented and was applied to acoustic emission data, with a focus on identifying and understanding damage processes, by considering the chronology of acoustic emission caused by damage during loading. The integration of optical fibres during the manufacture of flat and tubular composites allowed for monitoring of the strain gradients and residual strains caused by braiding, resin infusion, and curing, and global-local strain monitoring during mechanical loading of the specimens. Application of a simple damage index based on guided waves enabled detection of impact damage and damage accumulation during loading, while the experimental measurement of group velocities enabled estimation of acoustic emission source locations during loading. Analysis of acoustic emission data through clustering enabled identification of phases related to damage progression, which enabled data clusters to be linked to different damage mechanisms. The techniques applied in this work have been shown to complement each other without increasing the computational cost, making them suitable for application in industries such as oil and gas for continuous and periodic monitoring of composites. This has the potential to revolutionise the use of structural health monitoring as a tool for improving confidence and reliability, as well as informing and optimising the associated design and manufacturing processes.