The aim of this paper is to evaluate how different volatility estimation techniques impact the quality of pricing option contracts. The theoretical part explains option pricing, qualitative and quantitative parameters of the Black Scholes model, and implied volatility features. The pricing performance of the Black Scholes model with historical volatilities and of the ad hoc Black Scholes model with implied volatilities are assessed with Matlab, using a real option dataset consisting of S&P 500 call options. Moreover, the specification of the regression structure used in the ad hoc Black Scholes model to estimate volatility is analysed. It is shown that the absolute smile regression structure using strike price, time to maturity and their com- bination as independent variables for one-day ahead out of sample pricing is the most accurate technique for pricing options out of all the methods considered.