Progress is reported in an emerging non-contact strain sensing technology based on optical properties of single-walled carbon nanotubes (SWCNTs). In this strain-sensing smart skin ("S 4 ") method, nanotubes are dilutely embedded in a thin polymer film applied to a substrate of interest. Subsequent strain in the substrate is transferred to the nanotubes, causing systematic spectral shifts in their characteristic short-wave infrared fluorescence peaks. A small diode laser excites a spot on the coated surface, and the resulting emission is captured and spectrally analyzed to deduce local strain. To advance performance of the method, we prepare S 4 films with structurally selected SWCNTs. These give less congested emission spectra that can be analyzed precisely. However, quenching interactions with the polymer host reduce SWCNT emission intensity by an order of magnitude. The instrumentation that captures SWCNT fluorescence has been made lighter and smaller for hand-held use or mounting onto a positioning mechanism that makes efficient automated strain scans of laboratory test specimens. Statistical analysis of large S 4 data sets exposes uncertainties in measurements at single positions plus spatial variations in deduced baseline strain levels. Future refinements to S 4 film formulation and processing should provide improved strain sensing performance suitable for industrial application. In structural health monitoring, routine inspection and maintenance are vital to ensure the safe operation and maximum service life of critical structures, including bridges, high-rise building supports, oil platforms, pipelines, pressure vessels, and airframes. Strain measured on the surface of any structure provides direct first-hand information about its health condition. Established strain sensor technologies, such as resistance strain gages and fiber Bragg grating (FBG) sensors, and some newer ones, including piezoelectric sensors, 1 MEMS doubleended tuning fork (DETF) strain gauges, 2 and nanotube films, 3 are all point-wise, unidirectional sensing methods that require physical connections to the sensor elements in order to obtain strain readings. They are therefore classified as contact strain sensing methods. By contrast, the focus of this report is a new technology for noncontact strain sensing.Some existing full-field non-contact optical methods have also been developed, such as interferometric techniques, 4 noninterferometric techniques, 5-7 and Raman spectroscopy. 8 The interferometric techniques, including holography, interferometry, speckle interferometry, etc., usually require a model of actual structure, tedious calculations to separate the values of principal stresses, and expensive equipment. Non-interferometric techniques like the grid method, 5 and digital image correlation (DIC) 6,7 suffer from some other limitations, such as the requirements for a random gray intensity distribution or speckle pattern distribution, heavy dependence on the quality of the imaging system, relatively low strain measurement accuracy...