Knowledge of results was presented in pedestal sight gunnery training by means of a buzzer operative during trials, or spoken evaluative statements at the conclusions of trials. Amount and "pattern" of knowledge of results made little difference. Buzzer introduction and removal were associated with sharp rises and drops in scores. attributed to a cue-occurrence versus cue-nonoccurrence discrimination; no such effects accompanied the spoken statements. After cessation of knowledge of results, buzzer groups showed greater improvement than all others, but this advantage tended to diminish with practice; groups trained with the spoken statements performed consistently better than control groups. When transfer to a second training device was required, the buzzer Ss were at a disadvantage. It was concluded that the buzzer Ss had learned to capitalize upon situational factors particular to the original training device, and use of the buzzer was therefore not recommended. Definitive results were lacking with respect to the transferability of improvement resulting from the spoken statements.