This paper presents an inverse method for retrieving (i) the true thermal conductivity, and (ii) the two‐band absorption coefficient of soda‐lime silicate glassmelts between 1100°C and 1550°C from measured steady‐state temperature profiles. This was achieved by combining (i) a forward method solving combined conductive and radiative heat transfer accounting for temperature‐dependent thermal conductivity and spectral absorption coefficient and (ii) an inverse method based on genetic algorithm (GA) optimization. Four glassmelt compositions from ultraclear to gray glasses with iron content ranging from 0.008 to 1.1 wt% were investigated. First, it was established that the steady‐state temperature in glassmelt can be predicted accurately by averaging the spectral absorption coefficient over two bands from 0 to 2.8 μm and 2.8 to 5.0 μm. The inverse method showed that the true thermal conductivity was independent of the iron content and given by kc(T) = 1.31 + 5.90 × 10−4T, where T is given in °C. In addition, the band absorption coefficient between 0 and 2.8 μm strongly increased with increasing iron content, while the band absorption coefficient between 2.8–5.0 μm was independent of iron content.