Magnesium batteries are batteries that utilize magnesium cations as charge carriers and possibly in the anode in electrochemical cells. Both non-rechargeable primary cell and rechargeable secondary cell chemistries have been investigated. Magnesium primary cell batteries have been commercialised and have found use as reserve and general use batteries.

Magnesium secondary cell batteries are an active research topic as a possible replacement or improvement over lithium-ion–based battery chemistries in certain applications. A significant advantage of magnesium cells is their use of a solid magnesium anode, offering energy density higher than lithium batteries. Insertion-type anodes ('magnesium ion') have been researched.

Primary magnesium cells have been developed since the early 20th century. In the anode, they take advantage of the low stability and high energy of magnesium metal, whose bonding is weaker by more than 250 kJ/mol compared to iron and most other transition metals, which bond strongly via their partially filled d-orbitals. A number of chemistries for reserve battery types have been studied, with cathode materials including silver chloride, copper(I) chloride, palladium(II) chloride, copper(I) iodide, copper(I) thiocyanate, manganese dioxide and air (oxygen). For example, a water-activated silver chloride/magnesium reserve battery became commercially available by 1943.

The magnesium dry battery type BA-4386 was fully commercialised, with costs per unit approaching that of zinc batteries. Compared to equivalent zinc-carbon cells they had greater capacity by volume, and longer shelf life. The BA-4386 was widely used by the US military from 1968 until ca.1984, when it was replaced by a lithium thionyl chloride battery.

A magnesium–air battery has a theoretical operating voltage of 3.1 V and energy density of 6.8 kWh/kg. General Electric produced a magnesium–air battery operating in neutral NaCl solution as early as the 1960s. The magnesium–air battery is a primary cell, but has the potential to be 'refuelable' by replacement of the anode and electrolyte. Some primary magnesium batteries find use as land-based backup systems as well as undersea power sources, using seawater as the electrolyte. The Mark 44 torpedo uses a water-activated magnesium battery.

Secondary magnesium ion batteries involve the reversible flux of Mg²⁺ ions. They are a candidate for improvement on lithium-ion battery technologies in certain applications. Magnesium has a theoretical energy density per unit mass under half that of lithium (18.8 MJ/kg (~2205 mAh/g) vs. 42.3 MJ/kg), but a volumetric energy density around 50% higher (32.731 GJ/m³ (3833 mAh/mL) vs. 22.569 GJ/m³ (2046 mAh/mL). Magnesium anodes do not exhibit dendrite formation, albeit only in certain nonaqueous solvents and at current densities below ca. 1 mA/cm². This allows magnesium metal to be used without an intercalation compound at the anode, thus raising the theoretical maximum relative volumetric energy density to around 5 times that of a graphite electrode. Modeling and cell analysis indicate that magnesium-based batteries may have a cost advantage due to magnesium's relative abundance and ease of mining.

Applications had been recognised by the 1990s based on V₂O₅, TiS₂, or Ti₂S₄ cathode materials and magnesium anodes. However, instabilities in the discharge state and uncertainties on the role of water in the electrolyte limited progress. In 2000, Israeli researchers reported dendrite-free Mg plating in AlCl₃-ether electrolytes with a fairy high (>2 V vs. Mg/Mg²⁺) anodic voltage stability limit. In that work, however, a low voltage (and somewhat expensive) anode material (chevrel-type Mo₆S₈) was used for Mg²⁺ intercalation. Despite research following that discovery, all attempts to develop a high-voltage Mg²⁺ intercalation anode for chloroaluminate (and related) electrolytes failed. Electrochemical Mg²⁺ intercalation into many solid materials is well known, for example from aqueous electrolytes. The problem is to find anode materials that show intercalation from the same solutions, which display reversible Mg metal plating.

In contrast to the Mg-metal batteries, Mg-ion batteries do not use a Mg-metal anode, but rather a solid material capable of intercalating Mg²⁺ ions. Such batteries usually use an aqueous or other polar electrolyte. A commercially viable/competitive market niche for Mg-ion batteries has not been identified.