World population is constantly increasing and it is necessary to have sufficient crop production. Monitoring crop growth and yield estimation are very important for the economic development of a nation. The prediction of crop yield has direct impact on national and international economies and play important role in the food management and food security. Deep learning gains importance on crop monitoring, crop type classification and crop yield estimation applications with the recent advances in image classification using deep Convolutional Neural Networks. Traditional crop yield prediction approaches based on remote sensing consist of classical Machine Learning methods such as Support Vector Machines and Decision Trees. Convolutional Neural Network (CNN] and Long-Short Term Memory Network (LSTM] are deep neural network models that are proposed for crop yield prediction recently. This study focused on soybean yield prediction of Lauderdale County, Alabama, USA using 3D CNN model that leverages the spatiotemporal features. The yield is provided from USDA NASS Quick Stat tool for years 2003-2016. The satellite data used is collected from NASA's MODIS land products surface reflectance, land surface temperature and land surface temperature via Google Earth Engine. The root mean squared error (RMSE] is used as the evaluation metric in order to be able to compare the results with other methods that generally uses RMSE as the evaluation metric.