We have presented a number of optical techniques used to characterize properties of developed tissue phantoms designed for simulating optical properties of human skin at around 930 nm. The OCT imaging is primarily used as a noninvasive diagnostic technique and is often the recipient technique of the phantoms for calibration or research purposes.45–47 Moreover, it can be used to estimate the phantom geometrical thickness and RI with a precision dependent on its axial resolution. The thicknesses of the produced phantoms were measured with the OCT with an accuracy of (in air). However, some studies show techniques which use OCT for the determination of the microscopic properties.21,48–50 In-depth scanning allows for investigation of internal inhomogeneities, which are mainly caused by trapped air bubbles and clustering of nanoparticles. No significant inhomogeneities were detected in the produced phantoms. The surface profile of the phantoms was determined with an accuracy with the use of the white-light profilometer. Production errors such as channels created by evacuating air bubbles during solidification of the phantoms result in large variations in phantom thickness. We have noticed that the magnitude of this effect drops with the increasing thickness of the phantoms. Therefore, the relative error in determination of thickness was greatest for the thinnest phantoms. The maximum relative error is 0.1%, 1.5%, 3%, and 20% for the 2-mm-, 1-mm-, -, and -thick phantoms, respectively. The RI was measured for multiple wavelengths, which allowed us to define the dispersion function with correlation coefficient. The RI at 930 nm was approximately for all phantoms, which is very similar to that of the real tissues.28 The spectrophotometric measurements consisted of transmittance, reflectance, and collimated transmittance measurements. Reflectance and transmittance were collected using integrating spheres, while the collimated transmittance was collected with the use of a collimating lens. Measurements were done from 2 s up to 8 s integration time when necessary under low-light conditions. Spectra were obtained from several sites over the central area of the phantoms and then averaged. Since the setup required rearrangement for each type of the measurement, acquiring data from the same spot for all three parameters was impossible and averaging over the middle area of the phantoms had to be done. The reflectance was measured using the comparison method consisting of a total of 2 calibration and 2 sample measurements. The strong dependence on the phantom thickness is evident in transmittance and reflectance measurements. In both of these cases, the introduction of absorption leads to a decrease of the measured values, which means that the observations are in agreement with the theoretical assumptions. The amount of light in collimated transmittance measurements is extremely low and, therefore, the most susceptible to noise. The results vary greatly with the phantoms’ thicknesses, even in spot-to-spot measurements over the same phantom. The IAD method was used to calculate the , , and g factor. The averaged (over all samples) reduced scattering coefficient for 930 nm equals and for nonabsorbing and absorbing phantoms, respectively. This values lie in close proximity with the values found in the literature for the reduced scattering coefficient of human skin.11–14 This proves the correctness of theoretical assumptions and the proper fabrication of the phantoms for the specified goal of mimicking the optical properties of human skin at 930 nm. The absorption coefficient increased from to for all phantoms except the -thick. These values are in agreement with the average absorption coefficients found in the literature data.11–14 The absorption coefficient for thick phantoms was overestimated by . This must be mainly due to the small size of the phantom, which causes light to escape from the edges, thus breaking the IAD assumption which was found to cause overestimations of the absorption coefficient.30,31 The anisotropy factor results depend greatly on phantom thickness. Since the anisotropy theoretically depends only on the particle parameters and their amount and size distribution in a medium,27,28 it should be constant regardless of the sample thickness. This leads to the conclusion that the calculated factors are based on erroneous measurements. Possible sources of this error are: (1) surface roughness (thickness difference) of the phantom surface between the measured spots; (2) uncertainty in measurements of thickness; (3) low signal intensity on the detector when measuring so that the noise introduces a large amount of error; and (4) detection of partially scattered light in unscattered transmission measurement, which probably had the most influence on incorrect estimation. A quite reasonable experimental value for the factor of 0.8 was obtained for thin samples due to less influence of multiple scattered light on collimated transmittance measurements in that case and a weak sensitivity of the reduced scattering coefficient to sample thickness. The consequence of the latter has been reported as causing the estimation of optical properties dependent on the thickness of the measured material.30,31 This was also evident in our study, however, the and dependences on thickness were mostly acceptable as an expected inaccuracy of estimation caused by the limited accuracy of the measurements.