A stochastic frontier production model was applied to estimate technical efficiency in a sample of Italian organic and conventional cereal farms. The main purpose was to assess which production technique revealed higher efficiency. Statistical tests on the pool sample model suggested that differences between the two cultivation methods were significant from a technological viewpoint. Separate analyses of two sub-samples (93 and 138 observations for organic and conventional farms, respectively) found that conventional farms were significantly more efficient than organic farms, with respect to their specific technology (0.892 vs. 0.825). This implies that organic (conventional) cereal farmers could increase their income to 99.19 €/ha (40.95 €/ha). Analysis also estimated that land was the technical input with the highest elasticity for both technologies. Furthermore, findings indicated that 63.7% of the differentials between observed and best-practice output was explained by technical inefficiency for the conventional group, while this value was close to unity for organic farms. Some policy implications can be drawn from these findings.