Diamond occurs as one of the most precious and the strongest naturally occurring materials. It has always remained in demand because of its rarity, light refracting properties that make it appealing, and the urge to have it inlaid in the jewelry. Unlike silver and gold, finalizing its price is not an easy job. Its valuation is a significant aspect of the diamond industry, affected by various features like carat weight, clarity, cut, and color. This research paper studies the usage of regression techniques in machine learning for prediction of diamond prices. The dataset consists of features namely carat, color, cut and clarity, along with some physical dimensions like depth, x, y, z and table, which are either categorical or continuous. Some categorical features have been encoded into continuous (numerical) values so as to facilitate efficient and easier prediction. By analyzing the performance of different regression algorithms and assessing the outcomes based on the existing literature, the paper provides understanding of the effectiveness of regression analysis in price prediction of diamonds, clarifying the effect of different features on price variations. The regression algorithms implemented aim to obtain lower values of errors like MSE and RMSE, and greater value of R2 score. Random forest turned out to be the most efficient algorithm among them that effectively predicts diamond prices with an R2 score of 0.982 and minimum errors. Linear regression and decision tree also gave good R2 scores of 0.906 and 0.966, respectively.