The Atlas LV-3B, Atlas D Mercury Launch Vehicle or Mercury-Atlas Launch Vehicle, was a human-rated expendable launch system used as part of the United States Project Mercury to send astronauts into low Earth orbit. Manufactured by Convair, it was derived from the SM-65D Atlas missile and was a member of the Atlas family of rockets. With the Atlas having been originally designed as a weapon system, testing and design changes were made to the missile to make it a safe and reliable launch vehicle. After the changes were made and approved, the US launched the LV-3B nine times, four of which had crewed Mercury spacecraft.

The Atlas LV-3B was a human-rated expendable launch system used as part of the United States Project Mercury to send astronauts into low Earth orbit. Manufactured by American aircraft manufacturing company Convair, it was derived from the SM-65D Atlas missile and was a member of the Atlas family of rockets. The Atlas D missile was the natural choice for Project Mercury, as it was the only launch vehicle in the US arsenal that could put the spacecraft into orbit and also had many flights from which to gather data.

The Atlas had been originally designed as a weapon system. Thus, its design and reliability did not need to necessarily be 100% perfect, which led to Atlas launches frequently ending in explosions. As such, significant steps had to be taken to human-rate the missile to make it safe and reliable, unless NASA wished to spend several years developing a dedicated launch vehicle for crewed programs or else wait for the next-generation Titan II ICBM to become operational. Atlas's stage-and-a-half configuration was seen as preferable to the two-stage Titan in that all engines were ignited at liftoff, making it easier to test for hardware problems during pre-launch checks.

Shortly after being chosen for the program in early 1959, the Mercury astronauts were taken to watch the second D-series Atlas test, which exploded a minute into launch. This was the fifth straight complete or partial Atlas failure, and the booster was at this point nowhere near reliable enough to carry a nuclear warhead or an uncrewed satellite, let alone a human passenger. Plans to human-rate Atlas were effectively still on the drawing board, and Convair estimated that 75% reliability would be achieved by early 1961 and 85% reliability by the end of the year. Despite the Atlas' developmental problems, NASA had the benefit of conducting Project Mercury simultaneously with the Atlas R&D program, which gave plenty of test flights to draw data from, as well as test modified equipment for Mercury.

Aside from the modifications described below, Convair set aside a separate assembly line dedicated to Mercury-Atlas vehicles which was staffed by personnel who received special orientation and training on the importance of the crewed space program and the need for as high a degree of top-quality workmanship as possible. Components used in the Mercury-Atlas vehicles were given thorough testing to ensure proper manufacturing quality and operating condition. In addition, components and subsystems with excessive operating hours, out-of-specification performance, and questionable inspection records would be rejected. All components approved for the Mercury program were earmarked and stored separately from hardware intended for other Atlas programs and special handling procedures were taken to protect them from damage. The factory inspection of Mercury vehicles was performed by Convair personnel specially chosen for their experience, familiarity with the Atlas hardware, and who had demonstrated a favorable disposition and work ethic.

Propulsion systems used for the Mercury vehicles would be limited to standard D-series Atlas models of the Rocketdyne MA-2 engines which had been tested and found to have performance parameters closely matching NASA's specifications. NASA decided that the best choice of engines would be units with roughly medium-tier performance. Engines with higher than average performance were not considered acceptable because it could not be determined exactly why a given set of engines performed the way it did, and so it was considered safest to use medium-performance ones.

For the most part, NASA preferred to stay conservative with the Mercury vehicles and avoid modifying them any more than necessary. Modifications to the Atlas would largely be limited to those that improved pilot safety, and the standard D-series Atlas configuration was to be retained as much as possible, so assorted enhancements made to the latest Atlas missiles would not be used. Various equipment and procedures used with Mercury vehicles, although outdated and often not the best or latest, were preferred because they were proven and well-understood. Any new equipment or hardware changes made to Mercury vehicles had to be flown on at least three Atlas R&D tests before NASA would approve them for use. Despite the conservatism and caution taken with the design of Mercury vehicles, a large number of changes did take place over the 4+1⁄2 years of the program from lessons learned, and the emphasis on quality control got tighter as time went along; the last two Mercury flights were given a level of testing and pre-flight inspection that was unheard of when Big Joe flew in 1959.

All launch vehicles would have to be complete and fully flight-ready at delivery to Cape Canaveral with no missing components or unscheduled modifications/upgrades. After delivery, a comprehensive inspection of the booster would be undertaken, and prior to launch, a flight review board would convene to approve each booster as flight-ready. The review board would conduct an overview of all pre-launch checks and hardware repairs/modifications. In addition, Atlas flights that had taken place over the past few months in both NASA and Air Force programs would be reviewed to make sure no failures occurred involving any components or procedures relevant to Project Mercury.