10 psychiatric inpatients with a diagnosis of depression were entered into a prospective study to investigate the repeated one-point method as a method of predicting steady-state concentrations for nortriptyline. Four males and 4 females completed all requirements of the protocol. Single plasma concentrations obtained after the first and second daily doses were found to be accurate predictors of the steady-state minimum plasma concentrations. The difference between the measured and the predicted steady-state minimum values ranged from 2.10 to 32.7 micrograms/L with a standard deviation 10.4 and a mean of 10.5 micrograms/L, assuming a normal distribution. The correlation coefficient of predicted versus measured concentrations was 0.946 with 90% confidence limits, ranging from 0.800 to 0.990. The repeated one-point method was found to be an accurate predictor of steady-state minimum plasma concentrations in patients receiving nortriptyline. This method should allow for individual adjustment of dose while minimising the time required to achieve therapeutic, non-toxic plasma concentrations.