This document presents a system for soil analysis and crop prediction using data mining techniques. The system measures soil parameters like pH, nitrogen, phosphorus and potassium using sensors. It then uses a decision tree algorithm to classify the soil and predict suitable crops. The pH value is used to estimate other nutrient values. The nutrient values and soil type are sent over WiFi to a server, which uses machine learning to predict crops and provide fertilizer recommendations to the farmer. The proposed system automates the soil testing process and aims to help farmers select optimal crops and increase agricultural yields.