The advent of spine research has gradually improvised using Artificial Intelligence and Machine learning techniques. Data mining had a greater influence on minimizing the complexities of medical diagnosis. The most concentrated aspects of medical diagnosis are feature selection and classifications. Accurate classification minimizes human intervention especially in handling most sensitive cases like vertebral column disorders. In the medical domain, the major controversy is distinguished from healthy and unhealthy spines. There were several existing methods are available for classifying the vertebral column disorders. But still, the need for advancement has grabbed attention among the research communities. In this paper, we proposed a random forest method for achieving enhanced classification results. The proposed mechanism consists of two phases and the first phases consist of random forest creation using N decision trees. The second phase is accurate prediction using the obtained tree structure from the first phase. The proposed method of efficiency is calculated based on accuracy, specificity, sensitivity, and F-measure. The experimental work is done using UCI medical dataset on MATLAB and WEKA tools. The observation is conducted between SPRINT [12], Ensemble Classifiers [11] with our proposed RF model. From the observation, it is proved the classification accuracy is quite improved with the implementation of our proposed random forest mechanism. The prediction achieved by random forest is 98 % which is more efficient in comparison to others especially on the terms of high accuracy and classifying speed.