The relationship between electrical activity, dopant solubility, and diffusivity was investigated as a function of the substrate temperature during implantation of Te, Cd, and Sn ions into GaAs. Implant doses of these species in the range 5×1012–5×1015 cm−2 were performed in the temperature range −196 to 400 °C, followed by either transient (950 °C, 5 s) or furnace (450–900 °C, 20 min) annealing. The redistribution after such annealing was found to depend on the implant temperature, and was always greatest for Cd followed by Sn and Te. The degree of electrical activation was in the same order, but there was essentially no correlation of electrical activity with dopant solubility. Te, for example, showed soluble fractions of ∼90% for a dose of 1015 cm−2 after annealing at 850 °C or higher, regardless of the initial implant temperature. By sharp contrast, the electrically active fraction under these conditions was in the range 0.8%–3.4%. There was also no apparent correlation of the degree of electrical activity with the presence of defects visible in transmission electron microscopy. The energy required to activate the implanted ions fell broadly into two categories: ‘‘low’’ values in the range ∼0.4–0.8 eV (which included Cd implanted or annealed under any condition, and elevated temperature implants of Sn and Te), and ‘‘high’’ values in the range 1.7–1.9 eV [which included implants of Sn and Te performed at −196 °C, or high dose (1015 cm−2) room‐temperature implants of these species].