Agriculture plays an important role for food security and financial stability. But, agricultural production is mostly affected by different factors such as weather condition, type of soil, and crop selection. To resolve this problem, a machine learning-based crop recommendation is proposed in this work. To find the suitable crop for a farm located in a geographical locality, the system takes prerecorded agricultural, weather, and soil-related information. The proposed system investigates the input data and recommends the most appropriate crop using different machine learning-based algorithms such as decision tree, random forest, and support vector machine. During the preprocessing phase missing values and duplicate values are identified from the dataset. The correlation between features is found to select the useful features. Finally different machine learning algorithms are trained and validated using the dataset. Among all different models, random forest has given the highest classification accuracy of 99.54%. The proposed system can be useful for farmers to finalize any crop before the farming. The system also can be used as an important tool to maximize the crop yield and avoid crop loss.