The effects of maximum, minimum, and slot length limits on the sizes and relative abundance of northern pike Esox lucius were evaluated in 22 Minnesota lakes. The regulations were implemented in 1989-1998 and lasted 9-15 years. As preregulation information was available back to the 1970s, the evaluation periods covered 21-37 years. Comparisons were made with reference populations from 47 ecologically similar lakes during the same extended period. Although the regulations did not achieve management objectives in every lake, the broader-scale, statewide finding was that they improved the size structure of northern pike populations but produced no consistent trends in relative abundance. The improvements were detected against the backdrop of reference populations that initially appeared to have similar sizes and relative abundances. Maximum length limits protecting fish over 20, 22, and 24 in produced significant long-term increases in the percentages of northern pike 24 in and longer and 30 in and longer compared with the reference populations. Lakes with 30-in minimum length limits had increased percentages of northern pike 20 in and longer, but the improvements did not carry over to fish 30 in and longer. A mix of slot length limits produced results that are more difficult to interpret but generally improved size structure. A metaanalysis incorporating all of the length regulations indicated that the changes in northern pike size structure in regulated lakes were very large for an ecological experiment. Length limits protected large northern pike, with the expectation that lower yields were an acceptable trade-off for producing larger fish for recreational fisheries. This study reveals the range and magnitude of responses we can reasonably expect from length limits as well as the substantial value of conserving large fish when the goal is improved population size structure. [ABSTRACT FROM AUTHOR]