Cancer diagnosis is typically delayed to the late stages of disease due to the asymptomatic nature of cancer in its early stages. Cancer screening offers the promise of early cancer detection, but most conventional diagnostic methods are invasive and remain ineffective at early detection. Breath analysis is, however, non-invasive and has the potential to detect cancer at an earlier stage by analyzing volatile biomarkers in exhaled breath. This paper summarizes breath sampling techniques and recent developments of various array-based sensor technologies for breath analysis. Significant advancements were made by a number of different research groups in the development of nanomaterial-based sensor arrays, and the ability to accurately distinguish cancer patients from healthy controls based on the volatile organic compounds (VOCs) in exhaled breath has been demonstrated. Optical sensors based on colorimetric sensor array technology are also discussed, where preliminary clinical studies suggest that metabolic VOC profiles could be used to accurately diagnose various forms of lung cancer. Recent studies have demonstrated the potential of using metabolic VOCs for cancer detection, but further standardization and validation is needed before breath analysis can be widely adopted as a clinically useful tool.