The fuel air ratio imbalance between individual cylinders can result in poor fuel economy and severe exhaust emissions. Individual cylinder fuel air ratio control is one of the important techniques used to improve fuel economy and reduce exhaust emission. California Air Resources Board (CARB) also has required automotive manufacturers to equip with on-board diagnosis system for cylinder fuel air ratio imbalance detection starting in 2011. However, one of the most challenging tasks for the individual cylinder fuel air ratio control and cylinder imbalance diagnosis is how to retrieve the cylinder fuel air ratio information effectively at low cost. This paper presents a novel and practical signal processing based fuel air ratio estimation method for individual cylinder fuel air ratio balance control and on-board fuel air ratio imbalance diagnosis. Based on temporal array signal processing techniques, an array of data samples from an oxygen sensor located in a confluence point of runners is fed into each cylinder's linear, non-linear, or neural network estimator to estimate its fuel air ratio. This method works with both wide range oxygen sensors and switching oxygen sensors. This paper presents in more detail the linear estimation method and the vehicle test results due to its advantages of good performance, low computational load, and easily automated calibration for production.