Various metal matrix composites (MMCs) are widely used in the automotive, aerospace and electrical industries due to their capability and flexibility in improving the mechanical, thermal and electrical properties of a component. However, current manufacturing technologies may suffer from insufficient process stability and reliability and inadequate economic efficiency and may not be able to satisfy the increasing demands placed on MMCs. Semi-solid powder processing (SPP), a technology that combines traditional powder metallurgy and semi-solid forming methods, has potential to produce MMCs with low cost and high efficiency. In this work, the analytical study and experimental investigation of SPP on the fabrication of MMCs were explored. An analytical model was developed to understand the deformation mechanism of the powder compact in the semi-solid state. The densification behavior of the Al6061 and SiC powder mixtures was investigated with different liquid fractions and SiC volume fractions. The limits of SPP were analyzed in terms of reinforcement phase loading and its impact on the composite microstructure. To explore adoption of new materials, carbon nanotube (CNT) was investigated as a reinforcing material in aluminum matrix using SPP. The process was successfully modeled for the mono-phase powder (Al6061) compaction and the density and density distribution were predicted. The deformation mechanism at low and high liquid fractions was discussed. In addition, the compaction behavior of the ceramic-metal powder mixture was understood, and the SiC loading limit was identified by parametric study. For the fabrication of CNT reinforced Al6061 composite, the mechanical alloying of Al6061-CNT powders was first investigated. A mathematical model was developed to predict the CNT length change during the mechanical alloying process. The effects of mechanical alloying time and processing temperature during SPP were studied on the mechanical, microstructural and compositional properties of the Al6061-CNT composites. A shear lag model was applied to predict the mechanical property (hardness) of the composite. This work demonstrated the promising potential of SPP in the fabrication of particle/fiber (nanotube) reinforced MMCs.