The study offers data on the technical efficiency (TE) and productivity growth of Polish crop farms. The data was obtained using Stochastic Frontier Models (SFM) and Data Envelopment Analysis (DEA). The application of these alternative approaches makes it possible to provide new information about production processes and indicates the consequences of using each method in efficiency and productivity analysis. The average TE scores obtained from SFM and DEA are 0.63 and 0.52 respectively. An analysis of exogenous factors affecting efficiency revealed that the size of agricultural area utilised has the strongest impact on efficiency in the DEA, while subsidies for less favoured areas have the strongest impact on efficiency in the SFM. In both methods, production elasticity with respect to materials was the highest, followed by elasticity with respect to labour. Moreover, both approaches indicate a productivity decline in the analysed period, though the causes of the decrease are different. The results obtained from SFM indicate that the TFP decline is attributed mainly to a decrease in technical efficiency not compensated by strong technical progress and small but positive scale growth. The opposite result was obtained using DEA, which indicates that the TFP decline was mainly caused by technical regression accompanied by small but positive scale growth.