This paper deals with estimation of technical change through changes in input factor efficiency/productivity. The main advantage of this factor augmenting (FA) approach, unlike the generic time trend models of technical change, is that one can measure input-specific productivity, change in input productivity, and the contribution of each input to overall (aggregate) technical change. In general, one can obtain more information about the nature of technical change from the FA approach compared to the popular time trend models. We model FA technical change in both primal and dual setups. In the dual approach we consider both cost and profit functions in the single and multiple output cases. Output efficiency factors are also added in the multiple output FA cost function model. Time series and panel (state level) data from U.S. agriculture are used to estimate several FA cost function models. The standard time trend models are also estimated using both the data sets. A translog cost system is used for all the specifications. Finally, results from alternative models are analyzed.