Recent computational innovations in horticulture are geared towards giving farmers reliable tools with which to predict disease before it happens. Such disease warning systems should enable farmers to take preventive measures before their crops are infected. One of the main inputs for plant disease warning systems is leaf wetness. Leaf wetness refers to the presence of water drops on leaf surface, and is caused by rainfall, dew, or guttation. Leaf wetness can be determined either by measurement or estimation using meteorological and other plant variables. Because leaf wetness or leaf wetness duration is so critical to the onset of many common fungal diseases in grapes there is a need, especially in wine growing countries like New Zealand, to find the most accurate way to determine leaf wetness. Leaf wetness sensors are not standard even in vineyards where sensors are deployed and these sensors are known to be costly to maintain and less than reliable. Leaf wetness measurements are taken using sensors that are placed in crop canopy and the data logged usually for later use rather than real time processing. There are different types of leaf wetness sensors with no universally accepted measurement standard. This thesis presents a comparative analysis of various sensors that are commercially available. The sensors used in the experiment have different sizes, shapes, and working principles. Visual observations were made and a sensor response time test was performed to evaluate sensors' performance. Using a dielectric constant based sensor to measure leaf wetness was shown to be more effective than resistance-based sensors. Paint application to the sensor also proved to increase sensitivity in larger sized sensors. The rapid development and varied nature of these sensors have contributed to a lack of standardisation and the lack of a single accepted protocol for the use of sensors. An alternative to the use of sensors is the simulation or modelling of leaf surface wetness. Simulation enables surface wetness to be estimated from historical, forecast weather data, or both, rather than from monitoring and measurement using in-field leaf wetness sensors. The primary focus of this thesis was to develop and evaluate a novel Adaptive Neuro-Fuzzy Inference Systems (ANFIS) as an approach to modelling leaf wetness duration in vineyards. A comparative analysis of ANFIS with Classification and Regression Tree/Stepwise Linear Discriminant (CART), Number of Hours Relative Humidity Greater than 90% (NHRH>90%) model, Fuzzy Logic System (FLS) model, an Artificial Neural Network (ANN), the Penman-Monteith (P-M) model, and the Surface Wetness Energy Balance (SWEB) model is presented. The ANFIS model was found to have a higher accuracy than other models investigated. These findings indicate that ANFIS is a useful and practical solution for estimating leaf wetness duration using meteorological predictor variables.