The number of patients undergoing common surgical procedures varies widely across England because of funding restrictions, figures show.
Local NHS rationing of hernia repair, hip and knee replacements, cataracts and varicose vein surgeries has led to a "postcode lottery", say researchers.
Writing in the Lancet, they warned such policies could be "storing up problems for the future".
The Department of Health said access to services should not be decided on cost.
Suitability for surgery should be judged by clinical experts on the basis of individual need, a spokesman added.
There has been much anecdotal evidence about rationing of some surgical procedures on the NHS, but this has tended to focus on the number of policies in place rather than details on how this has affected patients being operated on.
And while there had been much debate about the "clinical value" of some elective surgical procedures, there was poor consensus on which treatments should be restricted to save costs, the Imperial College London team said.
A comparison of hospital data with primary care trust (PCT) policies on rationing of surgery showed a significant difference in the number of operations being done depending on local rules.
For cataract surgery, PCTs with rationing in place were admitting about 48% fewer patients than those with no such policy in the first year restrictions were introduced, the figures showed.
And for knee replacements there was at least a 20% difference in the number of patients having the operation, depending on whether there were restrictions in place, they found.
When rationing for hernia surgery was introduced in some areas in 2006-07, there were 59% fewer patients being operated on in some PCTs, compared with those with no policy.
The researchers said this gap had since narrowed but in 2010-11 there had still been a 15% deficit, which equated to 64 fewer operations per 100,000 people per year in areas with restricted access.
Significant variations had been seen in access for almost every year looked at since rationing policies had been in place, they said.
The type and number of procedures rationed by PCTs varied considerably and only 17 of 119 respondents had no policy for all five of the procedures looked at.
Study author Steve Beales, a health informatics and policy analyst, said the study did not address whether rationing itself was "good or bad" and there were legitimate arguments on both sides.
"But variation is a bad thing and there does need to be national guidance on this.
"NICE [ The National Institute for Health and Clinical Excellence] would be best placed to do that," he said.
Although evidence was not yet available on the wider consequences of local rationing of surgeries, it did raise the question of whether England was "storing up problems for the future", he added.
And he said with a greater number of Clinical Commissioning Groups taking over from existing PCTs in April, "it is entirely possible this will lead to greater variation".
A Department of Health representative said: "We have already written to the NHS to clearly set out that access to services should not be restricted on the basis of cost - it is wrong and compromises patient care.
"Decisions on treatments, including suitability for surgery, should be made by clinical experts taking the needs of each patient into account."