A concept for direct fusion drive based on centrifugal mirror (CM) confinement of thermonuclear plasmas (DFD-CM) is described. In centrifugal mirrors, electric and magnetic fields are combined to confine the plasma within a rapidly rotating annulus of burning plasma fixed between two mirror magnets. High energy fusion products leave the reactor core at a rate determined by the velocity of plasma rotation and the strength of the mirrors. Those departing through the jet-side mirror deposit their energy in a “warm” plasma that then expands through a magnetic nozzle to deliver jet power in the 100–1000 kW range. Fusion products departing through the power-side mirror are converted to electricity to power the reactor. Moderate thrusts at attractive specific impulses (10,000+ seconds) are possible. Findings are presented on CM reactor dynamics in propulsion applications, to include new insights into the relationship between mirror and centrifugal components of plasma confinement. Additionally, analysis will be presented on reactor operability limits and characterization of viable configurations based on power density, technology constraints, and the ability to self-power. Ongoing research into the physics of the warm plasma will be discussed, to include description of improved fidelity estimates for fusion energy deposition. Finally, considerations for Alfvén’s frozen-in theorem relative to fusion plasmas and magnetic nozzle performance will be outlined.Viable space commerce and future space exploration will require advanced power and propulsion technologies capable of multi-MW power generation with high specific impulse, moderate thrust levels, and low system specific mass. DFD could make a round trip to Mars in 3 months, and to Saturn, in less than 3 years. A DFD-CM powered mission could deliver an orbiter and 4 atmospheric probes to Neptune in four years. A second-generation Interstellar Probe powered by DFD-CM could reach the edge of the heliosphere (1000 astronomical units) in 25 years.