||Video encoder complexity reduction is a very active research area in the past 20 years. Many researchers have developed fast algorithms to reduce the com-putation complexity of video encoder in every aspects. In this thesis, we inves-tigated the reduction of video encoder computational complexity by exploiting relationship between different modules in the video encoder. We will explain the correlations between each modules and three algorithms are proposed to tackle the video encoder complexity problem. The first algorithm proposed to embed the quantization into discrete cosine transform (DCT) such that the DCT output do not need to be explicitly quantized at run time. The algorithm is called "New Quantized Discrete Cosine Transform" (NQDCT). Simulation results showed that the proposed algorithm can achieve a noticeable computational reduction with negligible visual quality degradation. The second proposed algorithm is the "Zero Value Quantized Discrete Cosine Transform" (ZQDCT). We derived the relationship between Sum Absolute Dif-ference (SAD) and quantized coefficients. Based on the value of the SAD, we can accurately predict the zero value quantized DCT coefficient. If the coeffi-cient is predicted to be zero, the computations on DCT, quantization, inverse-quantization and inverse-DCT can be skipped. Simulation results showed that a large reduction in computational complexity can be achieved without any loss in visual quality. Based on ZQDCT and QDCT, we improved the zero-valued coefficients predic-tion such that it is performed at an earlier stage, that allows more computations to be saved. The proposed algorithm is called "1D-Zero Quantized Discrete Co-sine Transform" (1DZQDCT). 1DZQDCT has higher threshold than ZQDCT at high bit rate which can help to detect more zero block, hence more computations can be skipped.