This attorney-disciplinary proceeding arose from a random audit of the trust funds of respondent, Michael A. Konopka. As a result of the audit, the Office of Attorney Ethics (OAE) filed a complaint against respondent charging him with misappropriation of clients’ funds, as well as failure to maintain required records and failure to safeguard clients’ funds. More specifically, a formal six-count ethics complaint filed in January 1988 against respondent charged him with violations of the Rules of Professional Conduct, namely, the failure to maintain trust-account records, commingling personal and client funds, failure to safeguard client funds, and knowing misappropriation of client funds, contrary to Rules of Professional Conduct 1.15 and 8.4.
Respondent entered a complicated family arrangement whereby he agreed with his parents that he would keep current the payments on two mortgages that covered his parents’ homestead and a two-family house that his soon-to-be-divorced sister owned with her husband. In brief, his parents did not want the family to lose the two-family income property. They mortgaged their own home for $80,000 and turned the funds over to respondent to clean up the arrearages on the first mortgage on the sister’s home and other obligations and to buy out the sister’s husband. In exchange for living in a unit of the two-family house, respondent was to collect rent from the other tenant and pay the mortgages and other bills related to the properties with those funds and his own funds.
Respondent maintained a ledger sheet entitled “Konopka, Paul and Eva, to Spencer Savings and Loan,” which, according to the Disciplinary Review Board (DRB), documented a client *227trust fund that respondent had established for his parents. The audit of respondent’s trust account, covering approximately a three-year period, revealed that respondent had made disbursements from the Konopka account for the Konopka properties in excess of the deposits in the account. Moreover, on many occasions, there were deficits in the accounts of respondent’s other clients at the same time that his disbursements exceeded his deposits in the Konopka account. As found by the DRB,
from August 20, 1982 through September 16, 1985, respondent regularly made payments related to his parents’ property that exceeded the amount on deposit [and on] at least 26 occasions during that three-year period, the Konopka account reflected a negative balance, in amounts ranging from $1,461.19 on August 20, 1982 to $6,374.37 on September 25, 1984. These negative balances resulted in the invasion of clients’ funds in the Edone, Wiegand, Armagost, Koceski, and Arslan matters.
The key issue is whether the trust-fund shortage was the result of knowing misappropriation or the product of inadvertent error or gross neglect in respondent’s handling of the funds in his accounts. The question is critical because a knowing misappropriation almost invariably calls for disbarment. In re Wilson, 81 N.J. 451, 409 A.2d 1153 (1979). Respondent denied any such knowledge.
Both the District Ethics Committee (DEC) and the DRB found that a knowing misappropriation of clients’ funds had occurred. We disagree. Based on our independent review of the record we find that the evidence of respondent’s conduct falls short of establishing clearly and convincingly the knowing misappropriation that the Wilson sanction seeks to deter.
I
The theory of the OAE’s case was that the imbalance in respondent’s accounts clearly established the Wilson violation. To summarize the disciplinary counsel’s opening remarks before the DEC:
And the allegations are simply this: That Mr. Konopka, the respondent here, kept bad books.
*228But we allege that it doesn’t just — it wasn’t simply * * * that he failed to maintain his client ledger sheets properly and didn’t have a running balance, but rather that his failure to maintain records impacted upon other clients’ funds that were in his trust account.
The bottom line though, was that when Mr. Prihoda [the OAE’s auditor] looked over the attachments — the so-called Konopka ledger sheets, and made up a running balance, he found that the account was out of trust on many occasions in the four year period that he made his analysis.
* * * And logic would tell us that if that account was out of sync, if there were more money disbursed from that account than had actually been received in it, then it’s likely that other client funds were used to make — you know for these excess disbursements. And that’s, indeed, what the proofs will show
Although there is no question that the entrusted funds did not remain intact, there is a genuine question whether respondent knowingly misappropriated those funds.
In a long series of cases, we have emphasized the need for clear and convincing proof that a knowing misappropriation has occurred. In re Simeone, 108 N.J. 515, 521-22, 531 A.2d 729 (1987), summarizes the principles that we apply in determining whether the case is more than one of “shoddy bookkeeping” and instead rises to the level of one of knowing misappropriation. We noted there that in a case such as In re Orlando, 104 N.J. 344, 517 A.2d 139 (1986), repeated and frequent instances of being out of trust did not necessarily add up to a knowing misappropriation. Those principles have been applied uniformly. See In re Librizzi, 117 N.J. 481, 569 A.2d 257 (1990); In re Gallo, 117 N.J. 365, 568 A.2d 522 (1989); In re Johnson, 105 N.J. 249, 520 A.2d 3 (1987).
At the same time, we have not retreated one bit from the principle that knowing misappropriation, when shown by clear and convincing evidence, will warrant the Wilson sanction of disbarment. See In re Sommers, 114 N.J. 209, 553 A.2d 789 (1989); In re Warhaftig, 106 N.J. 529, 524 A.2d 398 (1987).
To compare this case to In re Skevin, 104 N.J. 476, 517 A.2d 852 (1986), cert. denied, 481 U.S. 1028, 107 S.Ct. 1954, 95 L.Ed.2d 526 (1987), in which an attorney who knew that he had *229not yet received personal-injury settlement checks nonetheless drew advance fees from his account, is inappropriate. In this case there is no clear and convincing evidence that respondent was systematically taking advance fees or borrowing from one client’s fund to make up for a shortfall in another’s.
Were there such proof, we agree that respondent’s conduct would merit disbarment. The disciplinary panels, in reaching critical conclusions that respondent had knowingly misappropriated clients’ funds, relied on two pieces of evidence. We find that that evidence falls short of the necessary clear and convincing proof. The DRB summarized the two bases as follows:
In the Konopka ledger, in respondent’s own handwriting, the phrase “balance forward from page 22” appears, followed by the stated balance of $153.81. Two lines down, again in respondent’s handwriting, two $500.00 disbursements are listed, one on May 24, 1982 and the other on June 1, 1982. No deposit was made to cover these excessive disbursements until August 2, 1982, sixty days later, when $385.00 was deposited. This deposit decreased the shortage to $461.19.
********
In the Edone matter, respondent deposited $5,000.00 into his trust account on January 24, 1983. This $5,000.00 was subsequently paid over to Ms. Edone on February 15, 1983. However, as of February 13, 1983, the trust account reflected a balance of $3,867.39, creating a $1,132.61 shortage in the Edone funds. On the same day the $5,000.00 was disbursed to Mrs. Edone, respondent deposited $1,500.00 of his personal funds into the trust account, thus bringing the trust account total from $3,867.39 to $5,367.39, an amount sufficient to cover the disbursement.
We are unable to agree that those transactions establish clearly and convincingly the knowing misappropriation that is predicate to disbarment.
As noted, the DRB relied on what we believe are two telltale disbursements from the Konopka family account of $500 each in May and June 1982. The supposition is that because of the preexisting $153 “balance forward,” respondent was plainly aware that that account had a deficit, but blindly invaded other clients’ funds for the purposes of the Konopka family account. But the strongest criticism of respondent found in the report of Mr. Prihoda, the OAE’s auditor, was that respondent had never reconciled his accounts. The auditor indicated: “One column *230[of the ledger sheets] accounted for receipts and the other column accounted for disbursements. There was no third column to show the current balance of the ledger accounts at any given time.” The OAE did not prove exactly how or when Konopka made the notation of a running balance at the top of the ledger sheet. Mr. Prihoda did not testify in these proceedings. The OAE simply relied on the submission of the reports. When questioned on the knowledge implied by the recording of disbursements immediately following the inadequate balance, respondent replied, “The checks probably were written, and then I entered them in the book.” There is simply no proof of when Konopka made the “balance forward” entry in relation to the issuance of the checks.
With respect to the asserted imbalance in the Edone matter, the fact that respondent made a deposit of $1500 when he made the disbursement to that client does not establish that he knowingly misappropriated clients’ funds. In fact, we have criticized an attorney when he has not made up deficits in his account. See In re Brown, 102 N.J. 512, 509 A.2d 176 (1986) (respondent continually invaded trust funds of one client to pay another). In addition, we know from the record that after the OAE audit, respondent engaged Edone, the very client whom he is accused of betraying, to perform the accounting services necessary to reconcile respondent’s accounts. We believe it unlikely that Konopka would knowingly invade Edone’s funds and then hire that client as an accountant to audit respondent’s own books.
In short, there is no clear and convincing proof that respondent knew that he was invading clients’ funds when he made those disbursements. In fact, the first count of the OAE’s complaint showed that he had regularly used the trust account to receive personal fees. He later recognized that to be an inappropriate form of accounting for his professional fees. One such inadequacy was that respondent did not draw fees from his trust account for deposit to his attorney account, but rather drew directly on the trust account. Thus, at any given *231moment there may or may not have been uncollected fees in the trust account. In addition, respondent did not use his trust account for his own venal purposes. The two $500 disbursements were clearly related to the purposes of the Konopka family account. Compare In re Jacob, 95 N.J. 132, 137, 469 A.2d 498 (1984) (attorney misappropriated client funds “ ‘to sustain his dual life style____’ ”). There is no evidence of such corruption here.
II
Our opinion in this case should serve then to emphasize our refusal to disbar in knowing-misappropriation cases unless we are totally satisfied that the proofs are clear and convincing But despite our care in applying the Wilson rule, three members of this Court have questioned that rule in this case. We are not certain of the extent of their concerns. The concurrence agrees that “the Wilson rule is the right rule for the vast majority of knowing-misappropriation cases,” post at 241, 596 A.2d at 741, and that it should “continue to guide the Court in determining the discipline to be imposed.” Post at 259, 596 A.2d at 752.
To debate the position of the concurrence is perhaps premature, since the concurrence’s rule has not yet been applied in a knowing-misappropriation case. The concurrence suggests only that there should be exceptions to Wilson “under special circumstances,” post at 259, 596 A.2d at 752, exceptions “to accommodate unique circumstances,” post at 241, 596 A.2d at 741, that should be “sensible and limited exceptions,” post at 257, 596 A.2d at 751.
One clue to the scope of the concurrence’s possible modification of Wilson may be found in the example that the concurrence draws from this case. Although agreeing with the Court’s opinion that knowing misappropriation was not clearly and convincingly proved, the concurrence says that even if respondent was guilty of knowing misappropriation of his *232clients’ funds, he should not be disbarred. Post at 249-251, 258-259, 596 A.2d at 746-747, 751-752. The concurrence argues that “[ejven if this record is read to sustain the OAE’s contention that respondent knew that funds of other clients were being used to make mortgage payments for his parents and sister, there is serious doubt whether respondent understood that his method of operating the account was prohibited.” Post at 249, 596 A.2d at 746.
No more radical exception to Wilson could be designed. Stripped of euphemism (the concurrence calls it ignorance of “trust-account management,” post at 241, 596 A.2d at 741), essentially the concurrence says that a lawyer who misappropriates clients’ trust funds should not be disbarred if the lawyer did not understand that stealing those trust funds was wrong. Such a defense (never seriously pressed by respondent) would seem to support the contrary position: an attorney who does not know that to take a client’s funds is wrong ought no longer be practicing law.
That the concurrence would, unnecessarily, posit that defense even hypothetically as an example of an appropriate exception to Wilson suggests perhaps the depths of its dissatisfaction and the potential damage to the public’s protection against misappropriation if its exceptions become the law.
The main basis stated for the position of the concurrence lacks support in our decisions. The concurrence argues that our decisions following Wilson have been rigid and inflexible, suggesting that we have disbarred automatically, without thoughtful reflection and without consideration of the circumstances. In fact, all we have done is adhere consistently to Wilson’s central premise: that a lawyer who knowingly misappropriates a client’s money must be disbarred — not because the lawyer is necessarily irremediably corrupt and evil, but because to do otherwise risks the worst possible damage to the judiciary: loss of public confidence in the bar and in this Court.
*233This Court’s unwavering position has been that when you take a client’s funds without the client’s permission, it is wrong, totally wrong, whether or not you intend to return the money, whether or not indeed you do return it. Practically every attorney who knowingly misappropriates intends to return the money, and many do, usually by taking it from another client, and the theft is often not discovered until the attorney gets to the end of the road where there is that last client from whom he or she has taken money but for whom he or she can find no new funds to steal from other clients. This “lapping,” the use of one client’s account to make up for a shortage in another, repeated on and on, and sometimes undetected, is the, most common form of knowing misappropriation. Clients should not have to endure the agony of worrying whether the lawyer will indeed make good on his or her intent to return the client’s money, or whether the lawyer’s family will be able to, or whether the lawyer will be able to make up the money from other clients.
The other justifications for Wilson, including the fact that the offense is clear and well understood, both by the lawyer and by the public, and universally regarded as irredeemable, all set forth in that opinion, are both supportive of and secondary to its major premise, the maintenance of public confidence. From case to case this Court has examined and reexamined the rationale of Wilson, its implications and its application, and concluded that it is correct — not because we are inflexible, but because we have found it to be correct in case after case. We have attempted to be principled while continually monitoring the practical impact of those principles. And we have been satisfied that they achieve the goals of Wilson and that those goals justify the results of each case. And if we have not been “pragmatic,” post at 241, 596 A.2d at 741, it is because we have yet to see how pragmatism justifies discarding those principles and goals.
The concurrence says that our inflexibility is such that we routinely order disbarment in Wilson cases simply because the *234attorney’s books are not in order. “If knowing misappropriation equates with disbarment, then there is no need for the OAE, the DECs, the DRB, or this Court to look beyond bookkeeping entries in the trust ledger to decide misappropriation cases. Never mind that we are extinguishing a lawyer’s professional life permanently and irrevocably.” Post at 252-253, 596 A.2d at 748. The charge is unwarranted. We insist, in every Wilson case, on clear and convincing proof that the attorney knew he or she was misappropriating. Obviously, we consider the attorney’s records, if relevant, along with all other testimony, but if all we have is proof from the records or elsewhere that trust funds were invaded without proof that the lawyer intended it, knew it, and did it; there will be no disbarment, no matter how strong the suspicions are that flow from that proof. See supra at 228, 596 A.2d at 734, noting, for example, In re Librizzi, supra, 117 N.J. at 490-91, 569 A.2d 257; In re Gallo, supra, 117 N.J. at 371-73, 568 A.2d 522; and In re Simeone, supra, 108 N.J. at 521-23, 531 A.2d 729.
A recurring theme of the concurrence is the closeness of knowing-misappropriation cases and the unusual difficulty it discerns in reaching the factual conclusion that knowing misappropriation has or has not occurred, as if some abstruse principle of accounting practice were trapping unwary innocent attorneys. Some of the Court’s requirements for record-keeping and accounting may be difficult for some to understand, even to comply with, but that has nothing to do with knowing misappropriation. Even the most untutored, naive attorney is fully equipped to understand the knowing-misappropriation prohibition. All the lawyer needs to understand is what it means to steal or to “borrow” without permission. For instance, when a lawyer withdraws more money from a client’s trust account than is deposited, the lawyer needs no course in accounting to realize that if the withdrawal is not covered, the difference is coming from some other client’s account. It can happen, it sometimes does happen, that a lawyer simply was unaware of the trust invasion, or more accurately, usually, that clear and *235convincing proof of his or her knowledge has not been provided, but not because there is anything complicated or subtle involved. In re Orlando, supra, 104 N.J. 344, 517 A.2d 139.
The record, since Wilson, of misappropriation-disbarment decisions shows overwhelming unanimity. Of thirty-six opinions ordering disbarment, thirty-three were unanimous, two were six-to-one (including one opinion, In re Fleischer, 102 N.J. 440, 508 A.2d 1115 (1986), involving three attorneys), and one was four-to-three. In addition, eighteen attorneys were disbarred by order for knowing misappropriation, all of which disbarments were unanimous; and finally, since Wilson there have been 107 disbarments either through resignation or by consent arising from knowing misappropriation. Put differently, of the 163 orders disbarring attorneys for knowing misappropriation since Wilson, only five (including three issued in Fleischer) were not unanimous. Furthermore, we have been unanimous when rejecting a claim of knowing misappropriation, with the sole exception of one case (In re Perez, 104 N.J. 316, 517 A.2d 123 (1986)).
For more than ten years this Court has grappled with the problem of knowing misappropriation, the harshness of disbarment on occasion, and the justification for that harshness. After the most serious reflection, we have consistently adhered to Wilson. In re Steinhoff, 114 N.J. 268, 553 A.2d 1349 (1989) (drug dependency); In re Nitti, 110 N.J. 321, 541 A.2d 217 (1988) (compulsive gambling); In re Skevin, supra, 104 N.J. 476, 517 A.2d 852 (family hardship). It is an adherence based on principle and on policy, carefully thought out; but if the rule is wrong, the issue is too important to be dealt with by suggesting the rule requires undefined exceptions. The uncertainty, inevitably created, damages its deterrent impact.
The judiciary’s system in this state for disciplining the bar is strong and effective. Its rules and procedures are calculated both to educate and to deter. Our “bounced check” rule (requiring banks to inform the OAE whenever an attorney *236draws a check on a trust account and the check is returned for insufficient funds) is one of eleven in this country. Our rule of invariable disbarment in knowing-misappropriation cases has been followed by but few states.1 The result is an attorney disciplinary system second to none. Bumsted & Guttman, “New Jersey leads nation with stringent discipline,” Beyond The Law, Gannett News Service Special Report, pt. II at 12 (1986).
Improvement, of course, is possible, but suggestions have almost uniformly been in the direction of further strengthening — including most recently the recommendations of the ABA Commission on Evaluation of Disciplinary Enforcement. The calls for change in lawyer discipline do not include any demand by the bar for easing the rules of their enforcement. Most certainly, they do not include any request by the New Jersey bar for any change in the knowing-misappropriation rule. Quite the contrary, the New Jersey State Bar Association has strongly supported the invariable imposition of permanent disbarment for knowing misappropriation.
We wish to state without equivocation that it is the sense of this Committee that a lawyer who steals a client’s property has committed the gravest breach of trust and committed an assault on the reputation of every lawyer of our State. We recommend that such conduct, without exception, warrants immediate and permanent disbarment, criminal prosecution and punishment which is swift, sure and severe. * * * [T]he Committee * * * unreservedly endorses the holding of In re Wilson * * * and urges its strict and uniform application to all cases of lawyer theft of a client’s property.
[New Jersey State Bar Association, Report of Select Committee to Review Standards for Safeguarding Clients’ Property 1-2, 6 (1983).]
In this state, the organized bar wants more professionalism and a higher standard of ethics just as much as the public does. It would be ironic were this Court to give clients less protection from lawyers who steal client funds than the bar itself says is necessary.
*237When a lawyer takes clients’ money, the potential devastation inflicted on the clients and their families is incalculable and unpredictable. The lawyer who intends to return the money may never be able to do so. A lawyer who takes clients’ funds puts innocent families at risk of total ruin. That total ruin of the client results from total trust in the lawyer. The New Jersey State Bar Association’s voluntary creation in 1961 of a fund, among the first in the nation, now called the New Jersey Lawyers’ Fund for Client Protection and now administered by the Court, to compensate the victims of such misappropriation is clear testimony to the horrendous quality of the offense.
We are convinced that public confidence in the bar requires that a lawyer who knowingly takes a client’s money, no matter what the reason, should never again practice law. We are equally convinced that the public believes that once that rule is clear, lawyers will be much less likely to take their clients’ funds, no matter what the reason. This Court agrees with that judgment, both on the merits and on its implications concerning public confidence in the disciplinary system.
That from 1980, when Wilson was decided, to 1991, a period during which the attorney population of New Jersey doubled, the number of attorneys whose misappropriation of clients’ funds generated claims each year seems to have remained remarkably stable is of some significance.2
*238Given the apparent success of the Wilson rule, its acceptance by the public and the bar, the measured consistency of our application, and our continued concern for the efficacy of attorney discipline, no change is in order now.
Ill
In sum, respondent’s conduct did not amount to a knowing misappropriation subject to the Wilson rule. Respondent’s conduct differed from the conduct displayed in In re Fleischer, supra, 102 N.J. 440, 508 A.2d 1115 (one respondent acknowledged use of clients’ funds and intentionally designed a bookkeeping system that prevented respondents from knowing whether they were using clients’ funds); in In re Warhaftig, supra, 106 N.J. 529, 524 A.2d 398 (respondent acknowledged that with full knowledge he was violating the rules, he withdrew advance fees because of his cash-flow problem); and in In re Brown, supra, 102 N.J. 512, 509 A.2d 176 (respondent, for four years, knowingly invaded clients’ trust funds through a “lapping process” whereby on a routine basis he designated funds of one client to pay for another client’s needs and for office expenditures).
Respondent’s conduct is more analogous to the conduct of the respondents in two recent cases, In re Librizzi, supra, 117 N.J. 481, 569 A.2d 257, and In re Gallo, supra, 117 N.J. 365, 568 A.2d 522. In those eases the attorneys were guilty of flagrant record-keeping violations but not of intentional misappropriation. Although we find that respondent did not knowingly misappropriate funds, his conduct was extremely careless. His record-keeping was totally inadequate. As the concurring *239opinion amply demonstrates, respondent has an appalling lack of knowledge of the maintenance of separate trust balances for each client. The record clearly and convincingly establishes a violation of Rule of Professional Conduct 1.15, requiring the safekeeping of clients’ property and compliance with the record-keeping requirements of Rule 1:21-6.
Nevertheless, discipline is not imposed in order to punish the attorney but to protect the public against members of the bar who are unworthy of their trust.
The Edones (recall that respondent had hired John Edone, an accountant, to reconcile his books of account) wrote:
After we completed the audit, we discussed the shortcomings with Michael and instructed him with regard to the proper procedures to be followed in the maintenance of his trust account in light of our experience as accountants. Michael immediately implemented these instructions. If Michael’s account was out-of-trust, it was so as a result of mistake and/or negligence, but not intent.
The Wiegands wrote:
If Michael’s trust account was in an out-of-trust situation at any time, we can only say that it had to have occurred as a result of mistake and/or negligence. We are deeply aware of Michael’s commitment to the law.
One of the Armagosts, who had been a secretary of respondent and worked on the family trust, wrote:
The estate matter was not only a part of the office, it was a family matter to me. As a result, had I thought something to be amiss, or remiss, with regard thereto, I would have immediately brought it to Michael’s attention. The fact that I never had occasion to note any difficulties with this particular file, or any other matter, is indicative of the fact that any out of trust situation was not discovered or known by myself, [another secretary,] or Michael. The fact that the random audit, some years later, indicates that an out of trust situation existed can only reinforce the fact that it occurred through mistake or negligence. To say that Michael, or any of us, would put a matter out of trust intentionally, is not correct, since we all strove to be as careful as humanly possible to insure that the trust account was correct. I know for a fact that the balancing of bank statements and the checkbook were virtually the exclusive province of myself and [another secretary].
In determining the appropriate discipline, we consider the interests of the public, the bar, and the respondent.
We find that the following mitigating factors exist in favor of respondent: an otherwise unblemished record for almost twenty years of practice; the absence of any financial injury as a *240result of his ethical violations; and, on realizing his error, his quick and appropriate corrective measures, including retention of an accountant to handle his trust account. Respondent has assured us that he will have an accountant review his books and that his system of accounting will be in compliance with the rules. We believe that the discipline to be imposed in this case, then, should be closest to that imposed in In re Librizzi, supra, 117 N.J. 481, 569 A.2d 257. We therefore direct that respondent be suspended from the practice of law for a period of six months.
Respondent shall reimburse the Ethics Financial Committee for any administrative costs.
So ordered.
All of the members of the Court join in Parts I and III of this opinion. The Chief Justice and Justices CLIFFORD, HANDLER, and POLLOCK join in Part II of the opinion.
See In re Addams, 579 A.2d 190 (D.C.App.1990); In re Marks, 72 A.D.2d 399, 424 N.Y.S.2d 229 (1980); State of Oklahoma, ex rel. Oklahoma Bar Ass’n v. Raskin, 642 P.2d 262 (Okla.1982); Carter v. McCarthy, 544 A.2d 149 (R.I.1988).
Number of attorney claims paid by the New Jersey Lawyers’ Fund for Client Protection (formerly Clients' Security Fund) 1980 to 1990:
Year Number of Attorneys
1980 13
1981 12
1982 11
1983 9
1984 10
1985 9
1986 10
1987 12
*238Year Number of Attorneys
1988 6
1989 11
1990 12
When the attorney has generated multiple claims, sometimes paid by the Fund over several years, he or she is counted only once.