Soil organic carbon (SOC) is an important component of any soil which determines many of its properties. Nowadays, more and more attention is being paid to the SOC content determination in soils by not using the conventional, time-consuming and expensive technique, but by using colour image processing of soil samples. In this case, even the camera of modern smartphones can be used as an image source, making this technique very convenient and practical. However, it is important to maintain certain standardised conditions (light intensity, light incidence angle, etc.) when capturing the images of soil samples. In our opinion, it is best to use a regular scanner for this purpose, with subsequent image processing by graphic programs (e.g., Adobe Photoshop). To increase the reliability of the colour information obtained in this way, it is desired (if possible) to use a spectrograph or a monochromator in the subsequent calculation of reflection or brightness ratios. It is these two approaches that we have implemented in our work. As a result of the experiment, the values of brightness ratios (at 480, 650 and 750 nm wavelengths and integral brightness ratio), colour indicators (the hue, saturation and value [HSV], red, green and blue [RGB], CIE L*a*b* and cyan, magenta, yellow and key [CMYK] systems) and SOC content in Calcic Chernozem samples of the steppe zone of Ukraine were obtained. Using correlation analysis of the dataset, the existence of direct (r = 0.88–0.90) and inverse close relationships (r = −0.75–0.90) between SOC, values of brightness ratios and colour indicators of the soil samples were established. This allows us to develop predictive models. Statistical analysis showed that the models were significant when they were based on the values of brightness ratios at 650 nm wavelength, integral brightness ratio, V indicator in HSV system, R, G and B indicators in RGB system, C, M and K indicators in CMYK system and L* and b* indicators in L*a*b* system. The subsequent calculation of variation coefficients showed that the largest variability was observed in SOC indicators (CV = 0.72) and slightly less variability in the K index of CMYK system and brightness ratio values at 650 nm wavelength (CV = 0.67 and 0.53, respectively). Based on this, we believe that the models y = 0.0188 + 0.0535*x (x is the value of the K index in CMYK system) and y = 5.0716 – 3.2255*log10(x) (x is the value of brightness ratio at 650 nm wavelength) were the most statistically significant and promising parameters for determining SOC content (y in these equations) in Calcic Chernozem samples of the steppe zone of Ukraine.