Extinction, the repeated presentation of a conditional stimulus (CS) without the unconditional stimulus (US), is the standard paradigm to reduce conditional responding acquired by the repeated pairing of CS and US in acquisition. However, this reduction of conditional responding is prone to relapse. In rodent fear-conditioning, gradual extinction, the fading out of CS-US pairings during extinction, has been shown to reduce the return of fear. The current study replicated the gradual extinction procedure in human fear conditioning and assessed whether it reduced the return of fear due to ABA renewal and reacquisition. During extinction, one group received standard extinction, a second received gradual extinction (increasing the spacing of USs presented after the 1st, 3rd, 6th, 10th, and 15th CS+ trials), and a third received reversed extinction training (decreasing the spacing of USs presented after the 1st, 6th, 10th, 13th, and 15th CS+ trials). Larger renewal and faster reacquisition of differential electrodermal responses to CS+ and CS− were expected after standard and reversed extinction than after gradual extinction training. The results were inconclusive due to the failure to find extinction of differential electrodermal responses and US expectancy ratings in both gradual and reversed extinction groups. Despite successful extinction in group standard, renewal was only observed in US expectancy. Visualization of US expectancy ratings during extinction suggested that potential identification of the US presentation patterns during extinction in the gradual and reversed groups delayed extinction learning.