The influence of variations in atmospheric temperature and ozone profiles on the total ozone column (TOC) derived from a Brewer MKII spectrophotometer operating in Thessaloniki, Greece, is investigated using three different sets of ozone absorption cross-sections. The standard Brewer total ozone retrieval algorithm uses the Bass and Paur (1985) cross-sections without accounting for the temperature dependence of the ozone cross-sections which produces a seasonally dependent bias in the measured TOC. The magnitude of this temperature effect depends on the altitude where the bulk of the ozone absorption occurs. Radiosonde measurements for the period 2000 to 2010 combined with climatological ozone profiles were used to calculate the effective temperature of ozone absorption and investigate its effect on the retrieved ozone column. Three different ozone absorption cross-section spectra convolved with the instrument's slit function were used: those of Bass and Paur (hereafter BP), currently used in the standard Brewer retrieval algorithm; those of Brion, Daumont, and Malicet (Malicet et al., 1985; hereafter BDM); and the recently published set by Serdyuchenko et al. (2013 hereafter S13). The temperature dependence of the differential ozone absorption coefficient ranges between 0.09 and 0.13% per degree Celsius for BP, between −0.11 and −0.06% per degree Celsius for BDM, and between 0.018 to 0.022% per degree Celsius for S13, resulting in a seasonal bias in the derived TOC of up to 2%, 1.8%, and 0.4%, respectively. The temperature sensitivity of the differential ozone absorption coefficient for the Brewer spectrophotometer at Thessaloniki for the BP and BDM cross-sections is found to be within the range reported for other Brewer instruments in earlier studies, whereas the seasonal bias in TOC is minimized when using the new S13 cross-sections because of their small temperature dependence.