We conducted a retrospective evaluation of alternative management strategies for stocks in the New England groundfish complex that have had recent history of target catches being set above the level that defines overfishing. In many cases the original target catches were unsustainable and would have resulted in stock collapses if the target catch had been removed. We evaluated (i) alternative harvest control rules, (ii) whether or not to do projections, (iii) whether the inputs to the projections (starting abundance and future recruitments) should be modified, and (iv) whether the target catches should be smoothed to prevent large changes from year to year. The greatest reductions in target catches resulted when no projections were done and the target catch was fixed over the period between assessments. Large reductions in target catches also occurred when a downward adjustment was made to the starting abundance in the projections based on the retrospective pattern. Neither approach alone was sufficient to prevent overfishing for most stocks, but when used in conjunction with one another or with an alternative control rule that reduced the target harvest rate as biomass fell below the target, the magnitude and frequency of overfishing was greatly reduced for most stocks. Attempts to adjust recruitment based on perceived changes over time were also effective for a few stocks, while attempts to smooth the target catches over often resulted in increases in the target catches.