The land reform that has unfolded in Zimbabwe since 1980 used different models and had diverse consequences. Since the implementation of the fast tract land reform programme in 2000, Zimbabwe experienced heavy reduction in yield and output at farm level that led to a 70% shortfall in production to meet annual food requirements (Richardson, 2005). The economic crisis in Zimbabwe has been characterized by worsening food insecurity especially in the rural areas where harvests continue to be poor. In the beef sector, Zimbabwe has failed to meet its export quota to the EU. The shortfall in production to meet annual food requirements shows a very grim situation but do not tell us about the performance of resettled farmers who now occupy much of the productive land. The broad objective of the study was to determine and compare the production efficiency of resettled farmers in Zimbabwe across land reform models. In addition, the study determined land use intensity. The study was conducted in the Mashonaland Central Province of Zimbabwe mainly because a wide variety of field crops were grown by resettled farmers. The respondents were stratified into three groups. These were: beneficiaries of land reform before 2000 (resettle scheme), fast track A1 model and fast track A2 model. The three models differ on how they were implemented and supported and this might result in different efficiencies of the models. A total of 245 copies structured questionnaire were administered on the resettled farmers from June to September 2010. Descriptive statistics was applied to the basic characteristics of the sampled households. The effect of model of land reform, gender of the household head, marital status, age of the household head, education, household size, religion, dependence ratio, whether the farmer was fulltime or part-time in farming, experience of the farmers in farming at that environment, total land size owned by the farmers and soil type on revenue per hectare and land use rate were determined using the GLM procedure of SAS (2003). Significance differences between least-square group means were compared using the PDIFF test of SAS (2003). The relationship between Revenue and land utilization was examined using the Pearson‟s correlations analysis. Dependance between response variables that had an effect on either revenue per hectare or land utilization with all the other response variables was tested using the Chi-square test for dependance. To find the effect of arable land used and herd size on revenue per hectare and land use the RSREG Procedure of SAS (2003) was used. Input oriented DEA model under the assumption of constant return to scale was used to estimate efficiency in this study. To identify factors that influence efficiency, a Tobit model censored at zero was selected. The mean land use rate varied significantly (p<0.05) with the land reform model with A2 having highest land use rate of 67%. The A1 and old resettlement households had land use rates of 53% and 46%, respectively. Sex, marital status, age of the household head, education and household size significantly affected land use (P<0.05). Revenue per hectare was not affected by any the factors that were inputted in the model. Results from the DEA approach showed that A2 farmers (large land owners) had an average technical efficiency score of 0.839, while the lowest ranking model (A1) had an average score of 0.618. Small land holders (A1 and the old resettled farmers) are on average less cost-efficient than large land owners, with a score of 0.29 for the former compared with 0.45 for the latter. From the factors that were entered in the Tobit model, age of household head, excellent production knowledge and farmer status affected technical efficiency whereas allocative efficiency was only affected by good production knowledge, farm size, arable land owned and area under cultivation. Factors which affected economic efficiency of the resettled farmers are secondary education, household size, farm size, cultivated area and arable land owned. None of the included socio-economic variables has significant effects on the allocative and economic efficiency of the resettled farmers. Thus, the allocative and economic inefficiencies of the farmers might be accounted for by other natural and environmental factors which were not captured in the model.