Due to the recent increase in the availability of photometric time-series databases , the characterisation of low-mass eclipsing binaries for the study of their orbital and physical parameters is now possible in large samples and with good precision . We have identified and photometrically characterised a sample of 230 detached close-orbiting eclipsing binaries with low-mass main-sequence components in the Catalina Sky Survey . These low-mass stars have masses of M \leq 1.0 M _ { \odot } and orbital periods shorter than 2 days . The adopted method provided a robust estimate of stellar parameters ( as mass and fractional radius ) by using only light curves and photometric colours , since no spectroscopic information was available for these objects . A SDSS-2MASS ten-colour grid of composite synthetic and observed colours and the K-Nearest Neighbours method were employed to identify main-sequence stars and to estimate their effective temperatures , typically of T _ { eff } \leq 5720 K. Each light curve was modelled with the JKTEBOP code together with an asexual genetic algorithm to obtain the most coherent values for the fitted parameters . The present work provides an unprecedented number of homogeneous estimates of main stellar parameters in short-period low-mass binary systems . The distribution of the components of the investigated detached eclipsing binaries in the mass-radius diagram supports a trend of radius inflation on low-mass main-sequence stars . A relative increase of inflation for lower masses is also found and our results suggest that the secondaries are more inflated , i.e . they present larger radii than the primary components of same mass , when compared to stellar evolutionary models .