Variation in baseline controlling relations is suggested as one of the factors determining variability in stimulus equivalence outcomes. This study used single- comparison trials attempting to control such controlling relations. Four children learned AB, BC, and CD conditional discriminations, with 2 samples and 2 comparison stimuli. In Condition A, a mask always covered the S+ or the S-, each in 50% of the training trials, ensuring both sample- S+ and sample- S- controlling relations. In subsequent tests, children showed immediate equivalence formation. Condition B trained the same sequence of conditional discriminations with different stimuli, attempting to prevent sample- S+ control in the BC conditional discrimination. Two children did not show equivalence, whereas the other 2 did. Probes suggested that children who formed equivalence in Condition B acquired sample- S+ relations, even with training designed to prevent them. Results indicate that acquisition of both sample- S+ and sample- S- relations increases the probability of immediate equivalence formation. (Contains 1 table and 4 figures.)