Twenty years ago the answer was simple: a torn ACL in an active person was reconstructed. The randomised trials since then have made the question more interesting and the answer more individual, and it is worth understanding what they showed before deciding anything.
What the trials found
KANON (Sweden, 2010 and 2013) randomised 121 young, active adults with an acute tear to early reconstruction with rehabilitation, or to rehabilitation with the option of a later reconstruction if the knee needed it. At two years the two groups had improved by exactly the same amount, and at five years there was no difference in knee scores, in arthritis on X-ray or in the number of meniscal operations. Half of the rehabilitation-first group had chosen a reconstruction by five years; the other half never needed one.
COMPARE (Netherlands, 2021) repeated the design in 167 patients. At two years the early-surgery group scored slightly higher on a knee questionnaire, by five points on a 100-point scale, which the authors judged below the threshold that a patient would notice. Half of the rehabilitation-first group had surgery during follow-up. There were more re-ruptures and more opposite-knee tears in the early-surgery group, four and three against two and one.
ACL SNNAP (United Kingdom, 2022) asked a different question: what about the person who has already waited, whose knee is still unstable months after the injury? In 316 such patients, surgery beat further rehabilitation clearly, by about eight points on the same scale, and 41% of those allocated to rehabilitation needed an operation within eighteen months anyway.
The Cochrane review summarises the acute-injury evidence as showing no difference in patient-reported outcome at two or five years between early surgery and rehabilitation first, on low-quality evidence. A meta-analysis of twelve studies found surgery lowered the risk of a later meniscal operation and gave a more stable knee on examination, with no difference in arthritis.
What that means in practice
Put together, the trials say three things. Rehabilitation first is a legitimate strategy for an acute tear, and roughly half of people who take it never need an operation. Those who do need one lose nothing by having it later, provided the knee has not been giving way in the meantime. And a knee that is still unstable months after the injury should have surgery rather than more exercises.
So the decision turns on who is likely to cope without the ligament, and who is not.
Likely to cope. In the Delaware-Oslo cohort, about 54% of patients who chose rehabilitation had a successful outcome at two years, and the predictors were older age, female sex, a good single-leg hop test and good function on a questionnaire at the start. Among patients classified as potential copers after a structured strength and control program, about three quarters succeeded without surgery. The person who walks, cycles, swims, runs in straight lines and skis carefully, and whose knee feels trustworthy after a good rehabilitation program, often does not need a reconstruction.
Unlikely to cope, and better reconstructed:
- A knee that gives way, in sport or in daily life, despite rehabilitation. Each episode risks the meniscus and the cartilage.
- The athlete returning to pivoting sport: netball, the football codes, basketball, soccer, skiing. The younger and the more competitive, the stronger the case.
- A repairable meniscal tear. Meniscal repairs heal better alongside a reconstruction than alone, and the meniscus is worth more to the knee’s future than the ligament is.
- Instability that has already persisted for months, where the trial evidence favours surgery outright.
Why an unstable knee should not wait
An unstable knee damages itself slowly. In a meta-analysis of timing, reconstruction more than three months after injury roughly doubled the rate of medial meniscal tears found at surgery, and the rise in high-grade cartilage damage became predominant after twelve months. In a population study, patients treated without surgery had more than five times the rate of a later meniscal tear and about six times the rate of arthritis compared with those reconstructed early. These are the reasons a reconstruction is recommended promptly for the knee that keeps giving way, and the reasons it is not recommended at all for the knee that does not.
The first few weeks after injury are different. A reconstruction done on a hot, stiff, swollen knee is the recipe for a stiff knee afterwards, and the meta-analysis of trials shows no penalty for surgery within three weeks once the knee is quiet. Settling the knee first, with full extension and a working quadriceps, is the start of the program whichever way the decision goes.
Cross-bracing: the honest position
A Sydney group published a method in 2023 in which the knee is locked at 90 degrees of bend in a brace for four weeks and then straightened in stages, to bring the torn ends together. In their first 80 patients, 90% showed healing on MRI at three months, and those with the best-healed ligaments mostly returned to their sport; 14% re-injured the ligament, and two patients developed a clot in the leg before clot prevention was added to the protocol. It is a genuine advance in understanding, and the authors themselves call for trials before it changes practice.
A second series, published in 2026, applied the protocol to 40 athletes aged 16 to 40 in pivoting sports and found 70% had recurrent instability and 62% had a medial meniscal tear at two years, far worse than the reconstructed comparison group. The originating group disputes the methods of that study. Where that leaves a patient today: cross-bracing is an option to discuss for the right knee in a person who is not returning to pivoting sport, is unproven in trials, carries a clot risk, and is not something this practice recommends for athletes who intend to pivot again.
How the decision is made here
Not on the day of the injury, and not by rule. The knee is settled first. Its stability is examined once the swelling has gone, the MRI is reviewed for the meniscus and cartilage, and the conversation is about your sport, your age, your work and what you want the knee to do for the next thirty years. Where the answer is rehabilitation, it is a structured program with a review, and the option of surgery stays open. Where the answer is reconstruction, the next chapter covers the weeks before it.