Plasma hormones were estimated in 24 postmenopausal patients who had been castrated. Each was given a sub-cutaneous implant of either 100 mg or 50 mg of oestradiol, or 50 mg of oestradiol with 100 mg of testosterone, or 200 mg of testosterone. Plasma hormone estimations were repeated at two weeks, one month and then monthly for up to 12 months. Plasma follicle stimulating hormone (FSH) and luteinizing hormone (LH) concentrations were seen to fall at two weeks after all implants containing oestradiol. Plasma testosterone concentrations rose from a mean concentration of 1.0 nmol/l to 5.0 nmol/l and 6.7 nmol/l after implants of 100 mg and 200 mg of testosterone respectively. Implants containing oestradiol caused the pretreatment ratio of the concentrations of oestrone to oestradiol to change from 2:1 to 1:2. The implant of 100 mg of oestradiol caused the plasma oestradiol concentration to rise to a mean value of 602.3 pmol/l and those of oestrone to rise to 356.7 pmol/l. The more commonly used implants contain 50 mg of oestradiol and these caused the mean concentration of plasma oestradiol to rise to 346.7 pmol/l and oestrone to rise to 233.9 pmol/l. These values compare favourably with those attained after oral oestrogen therapy.