This study attempted to determine whether or not the unelicited GSR can be conditioned instrumentally, using a visual reinforcer. The Ss were 40 college students. All Ss had 2 min. of initial rest, during which a base response frequency and a response magnitude criterion were determined. A reinforcement period followed the rest period. Two groups received 8 min. or 16 min. of reinforcement following each unelicited GSR. Two additional groups were matched to the first two groups in the number of visual stimuli presented per minute of reinforcement, but the stimuli were not given contingent upon the occurrence of unelicited GSRs. Finally, a 4-min. period of extinction was given. A significant interaction between reinforcement conditions and minutes of reinforcement was found in the 16-min. groups but not in the 8-min. groups. In general, the groups receiving contingent lights tended to respond at a higher rate than the noncontingent groups. During extinction, the relative response frequency trends for the contingent groups'' curves tended to be opposite to those of the noncontingent groups (downward and upward, respectively), although this effect was not statistically significant. The results were interpreted as supporting the conclusion that the unelicited GSR can be conditioned instrumentally.