This paper presents an alternative approach in estimating the effect that technological knowledge has on the cost structure facing individual firms. The suggested method is applied to the Swedish kraft paper industry and relies on a comprehensive dataset for eight individual integrated kraft paper mills. The developed model is based on a two-step process. Step one, the estimation of a pure cost reduction index is derived using a flexible variable cost function which is utilising mill-specific dummy variables. In addition, this approach allows for an estimation of the pure unit cost reduction index that is devoid of scale and price effects which, if not dealt with can produce spurious results when estimating learning effects. Step two, is the estimation of a two factor dynamic learning curve model (2FDLC) using the estimated pure unit cost reduction index as dependent variable. The results suggest that the Swedish kraft paper industry has relatively little to gain in terms of cost reduction through a further technological learning. However, the method performed well, producing intuitive and statistically significant estimates indicating its usefulness in further analyses.