This study, titled “A Study on Predicting Stock Prices of Selected Indian IT and Pharmaceutical Companies,” aims to forecast stock prices using ARIMA and LSTM models. A total of 20 companies 10 from each sector were analyzed using five years of historical stock data sourced from Aditya Birla Money Limited. The forecasting period was set to 36 months. ARIMA was applied using the Box-Jenkins methodology in EViews, while LSTM was implemented in Python using TensorFlow and Keras. Results showed that LSTM better captured nonlinear trends, particularly in the pharmaceutical sector, while ARIMA performed well in more stable IT stocks. This comparative analysis highlights the strengths of each model and provides insights into investment potential across sectors, helping investors and analysts make more informed financial decisions.