The heterogeneous distribution of minerals in different rock types poses several challenges for assessing thermal conductivity, heat flow and temperature of sedimentary basins, especially when databases coming from the oil sector are the only source of information. The objective of this study was to develop a new methodology that uses well log data to better infer the thermal conductivity variations of sedimentary formations in order to evaluate heat flow and extrapolate temperature at depth. The methodology was applied to the St. Lawrence Lowlands basin, with constrains from the available oil and gas database not designed for geothermal exploration purposes. The main idea was to analyze quantitatively well log data with an inversion approach from limited reference wells and derive empirical relationships to calculate a thermal conductivity profile for each available well. Pressure and temperature corrections were then considered. These continuous logs of thermal conductivity were used to estimate the Earth's heat flux density using bottomhole temperatures and to extrapolate temperature at depth. A modified version of Poisson's equation was solved by the finite difference method for this purpose. The average temperature and its standard deviation obtained with this approach for the St. Lawrence Lowlands at 1000, 2500 and 5000 m depth is approximately 32 +/- 6.9 degrees C, 63 degrees C +/- 12.7 degrees C and 119 degrees C +/- 28.3 degrees C, respectively. Moreover, temperature ranges from 19 to 52 degrees C, 41 to 112 degrees C and 75 to 236 degrees C at 1000, 2500 and 5000 m depth, respectively.