The characteristics of knock metrics over a number of engine cycles can be an important reference for knock detection and control in internal combustion engines. In a spark-ignition (SI) engine, the stochastic nature of combustion knock has been shown to follow a log-normal distribution. However, this has been derived from experiments done with gasoline only and applicability of log-normal distribution to dual-fuel combustion knock has not been explored. Natural gas (NG) features itself with relatively high methane and octane numbers, hence, it can be applied as additional fuel blended with gasoline, which enables to increase knock limited spark advance thereby increasing thermal efficiency. To evaluate the efficacy and accuracy of log-normal distribution model for NG-gasoline blended fuel, a sweep of NG-gasoline fueling blending ratio by energy was conducted at two different speeds. The experimental tests were conducted on a single cylinder prototype SI engine equipped with two fuel systems: a Direct Injection system for gasoline and a Port Fuel Injection system for compressed NG. The knock suppression tests were conducted at 1500 RPM and 2000 RPM, 12.0 bar net indicated mean effective pressure wherein the engine was boosted using compressed air. The results from blending methane with gasoline show the log-normal distribution model provide a good fit to the measured distribution and captures the characteristics of the distribution. It is found that the knock distribution can be described as a same log-normal distribution model at same knock borderline, with changing blending ratio. In addition, the independency of log-normal distribution model to the knock distribution at different speed and blending ratio is examined and a high coefficient of determination (R2) is observed between them. Results of this investigation can be useful in improving combustion process in engines which have application in transportation, and stationary engines used for heat and power generation.