Post-implant annealing of InP and GaInAs is usually accomplished using thermal cycles of 10-30 minutes duration; this thesis reports the results of a systematic study of an alternative technique, 'rapid thermal annealing' (RTA), in which the implanted material is held at elevated temperatures for less than 180 seconds. Results were obtained using Hall-effect measurements, Rutherford backscattering (RBS), secondary ion mass spectrometry (SIMS), and photoluminescence (PL), amongst other methods. Iron-doped InP, implanted with magnesium, zinc or mercury was subjected to RTA and five different methods of protecting the InP surface were compared: the use of an indium-tin pseudobinary leads to tin incorporation and n-type surface layer formation above 700°C; encapsulating layers of phosphosilicate glass, SiO2, Si3N4 or a novel 'dual' layer of Si3N4/AlK may lead to p-type, semi-insulating or n-type behaviour. This is shown to be due to the indiffusion of silicon from these encapsulants into the implanted substrate; this indiffusion is enhanced by implantation damage. RTA in a phosphine ambient gives the best surface protection at elevated temperatures, but leads to substantial outdiffusion and loss of the implanted dopant. Electrically active p-type layers were successfully obtained from both zinc and mercury implants. GaInAs was implanted with beryllium, magnesium, zinc and mercury and electrically active p-type layers obtained following magnesium implantation; electrical results were, however, dominated by the quality of the starting material and not reproducible. 'Proximity' annealing under a GaAsP or GaAs cover piece gave adequate surface protection for GaInAs at annealing temperatures up to 800°C. The presence of an amorphous layer In InP and GalnAs is shown to be detrimental and the maximum amorphous thickness which can be fully regrown is found to be about 250 nm at 750 °C. It is suggested that solid phase epitaxy of thicker amorphous layers is inhibited by the local nucleation of microcrystallites within the remaining amorphous material and a model describing the regrowth of III-V compounds is presented. Substantial redistribution of the implanted dopant occurs during RTA of InP and GalnAs, the shape of dopant profiles is modified by both the residual damage present within the material and the form of surface protection employed. Several models of acceptor diffusion in iron-doped InP are compared and discussed.