We aimed to develop a more accurate transpiration model for cucumber (Cucumis sativus L.) plants to optimize irrigation and nutrient usage in soilless greenhouse cultivation. Accurate modeling of transpiration in greenhouse-grown cucumbers is crucial for effective cultivation practices. Existing models have limitations that hinder their applicability. Therefore, this research focused on refining the modeling approach to address these limitations. To achieve this, a comprehensive methodology was employed. The actual transpiration rates of three cucumber plants were measured using a load cell, enabling crop fresh weight changes to be calculated. The transpiration model was developed by making specific corrections to the formula derived from the Penman-Monteith equation. In addition, the study investigated the relationship between transpiration rate and solar radiation (Rad) and vapor pressure deficit (VPD), identifying a nonlinear association between these variables. The transpiration model was adjusted to account for these nonlinear relationships and compensate for Rad and VPD. Comparative analysis between the actual and estimated transpiration rates demonstrated that the developed cucumber transpiration model reduced overestimation by 23.69%. Furthermore, the model exhibited higher coefficients of determination and root mean square error (RMSE) values than existing models, suggesting its superior accuracy in predicting transpiration rates. Implementing the transpiration model-based irrigation method demonstrated the potential for ∼21% nutrient savings compared with conventional irrigation practices. This finding highlights the practical applications of the developed model—accounting for a nonlinearity of Rad and VPD—in optimizing irrigation practices for greenhouse cucumber cultivation.