These days information is crucial in numerous fields, for example, such as, medicine, science, and business, where databases are utilized effectively for data sharing. In any case, the databases face the risk of being pirated, stolen or abused, which may bring about a lot of security threats concerning proprietorship rights, data tempering and privacy protection. Watermarking is used to uphold proprietorship rights on shared relational databases. Numerous reversible watermarking strategies are proposed recently to protect rights of proprietors along with recovering original data. Most state-of-the-art methods modify the original data to a large extent, result in data quality degradation, and can’t accomplish the good balance between robustness against malicious attacks and data recovery. In this project, we propose a robust and reversible database watermarking strategy, Genetic Algorithm and Histogram Shifting Watermarking (GAHSW), for numerical relational data. The genetic algorithm is utilized to choose the best secrete key for grouping database, where the watermarking can be embedded with adjusted mutilation and limit. The histogram of the prediction error is moved to insert the watermark with good robustness. Experimental results exhibit the effectiveness of GAHSW and demonstrate that it outperforms state-of-the-art approaches as far as robustness against malcious attacks and preservation of data quality.