Knowledge of energy in crop production is essential for enhancing food production for the rapidly increasing population. Adequate attention has not been given to the analysis of energy inputs in crop production in Nigeria. Consequently, it is difficultto determine viable energy inputs in the production processesand plan for conservation. Therefore, there is a need for evaluation of energy consumption in cassava production in Kwara State, Nigeria.The aim of this study was to determine energy requirement in cassava production in the State, and hence develop model equations for predicting yield under different farming practices. Thespecificobjectives of the study were to:i) investigate energy inputs and output per hectare in cassava production; ii) determine energy use efficiency in cassava production; iii) investigate the effects of energy inputs on cassava productivity; and iv) investigate the relationship between cassava yield and different energy input sources. rnrnData from 175 farms were collected through site visits and physical measurements for 3 consecutive cropping seasons (2013 - 2016). The farms were classified into 3 groups based on size of farm and level of mechanization: small scale (1 to 5 ha), medium scale (6 to 15 ha) and large scale (16 to 50 ha). There were 92 farms in the small scale group while medium and large scale farms were 54 and 29 farms, respectively. Inputs used in cassava production (cassava setts, chemicals, diesel fuel, fertilizers, human labour and machinery) and output (cassava yield) were determined per hectare and then multiplied by the coefficient of energy equivalents. The data generated were statistically analyzed using R software. Cobb-Douglas production function and multiple linear regression analysis were applied to analyze the relationship between energy inputs and cassava yieldto develop yield predictive models.rnrnThe findings of the study were that:rni) cassava production consumed a total energy of 4904.87 MJ/ha, 36352.04 MJ/ha and 96 257.93MJ/ha while average energy output were 107,632 MJ/ha, 604,800 MJ/ha and 2,016, 000 MJ/ha, respectively in small, medium and large scale farms respectively;rnii) energy use efficiency for cassava production was 16.13, 16.69 and 20.94% for small, medium and large scale farms, respectively;rniii) average values of net energy used and energy productivity were respectively 100, 959 MJ/ha and 2.88 for small scale farms, 568, 560 MJ/ha and 2.98 for medium scale farms, and 1, 919, 742 MJ/ha and 3.74 for large scale farms; andrniv) indirect and non- renewable energies had higher impacts on cassava yield than direct and renewable energies in medium and large-scale farms. Reverse was the case in small scale farms; rnThe study concluded that input variables in cassava production are independent of each other. Three parametric models were developed in this study to predict cassava yield. Model I with the highest R2 value and no autocorrelation at 5% significance levelis recommended to scientists in order to boost cassava yield.Results of this study would be useful to cassava farmers in Kwara State and elsewhere with similar soil characteristics and operating conditions.