Almost every divorce ends in an agreement. Judges in Colorado try a small fraction of the dissolution cases filed; the rest are resolved by the spouses, usually with lawyers, usually late, and usually after each side has spent a great deal of money learning what the other side already knew. Mediation moves that agreement to the front of the process instead of the end. That is the whole idea, and it is simpler than most descriptions make it sound.
What follows is how divorce mediation actually works in Colorado, where it goes wrong, what the best long-term research says about it, and the question couples rarely ask about the person they are hiring: what has this mediator actually seen?
What Mediation Is, and What It Is Not
A mediator is a neutral. The mediator does not represent either spouse, does not decide anything, and cannot impose a result. The spouses decide; the mediator runs the process, explains the law to both of them in the same words, and turns what they agree on into a document a judge will sign.
Three Colorado statutes frame the process. Under C.R.S. § 13-22-307, mediation communications are confidential: they cannot be disclosed voluntarily or compelled in discovery, the mediator cannot be made to testify about them, and a communication disclosed in violation of the statute is inadmissible. The exceptions are narrow: a threat of harm, intent to commit a felony, a statutory reporting duty, and a claim against the mediator. Under C.R.S. § 13-22-308, a mediated agreement that is reduced to writing and signed can be presented to the court and, once approved, enforced as an order. And under C.R.S. § 13-22-311, a court may refer a contested case to mediation, but not over the objection of a party who reports abuse. Many Denver-area judicial districts will not set a contested permanent orders hearing until the parties have at least tried.
The separation agreement that comes out of mediation is then reviewed by the court under C.R.S. § 14-10-112: the financial terms are binding unless unconscionable, and the children's terms are reviewed for the children's best interests. In most mediated cases the decree is entered on the papers, without a hearing, once the ninety-one-day waiting period has run.
What mediation is not: it is not therapy, it is not arbitration, and it is not "splitting the difference." A mediator who simply averages two positions rewards the spouse who asked for the most. The standard is not the midpoint. The standard is what a Colorado judge would do.
How It Works, Start to Finish
In my practice the process runs in nine steps, and most couples are through them in one to two months.
- Both spouses contact the office together. One request, both names. This is not a formality. If one spouse calls me alone about the divorce, I can be that spouse's lawyer or nobody's; I can no longer be the mediator, because a mediator who has heard one side first is not neutral.
- Conflicts check, and the rule attaches. If either spouse has consulted me about the divorce, I cannot mediate. If the check clears, the reverse rule attaches permanently: I will never represent either of them in the matter, whether or not the mediation succeeds. Colo. RPC 1.12 and 2.4 require a version of this; my own rule is stricter.
- The first session, both spouses present. Ninety minutes or so. Every concern is heard, the law on each issue is explained, and I screen for anything that makes mediation inappropriate.
- One flat fee for the whole mediation, quoted in the room. No hourly meter. A fee that grows with every session gives the mediator the wrong incentive.
- The agreement to mediate, and full disclosure. Both spouses commit in writing to disclose every asset, debt, and source of income, and exchange sworn financial statements and the documents the court would require before anyone negotiates.
- Joint sessions, four to six of them. Children first: decision-making, the schedule in overnights, holidays, exchanges, the child support worksheet. Then the money: the house, retirement accounts, the business, debts, maintenance, taxes. Individual meetings only at the mediator's discretion, with both spouses knowing they are happening.
- The complete agreement. A separation agreement covering every asset, debt, and support term; the parenting plan on the court's form; the child support worksheet; and the transfer documents that follow. Not a term sheet for someone else to finish.
- Independent review. Each spouse takes the complete agreement to a lawyer of their own before signing. That lawyer reads it with one client's interests in mind. I will not hold the signing session without it.
- The signing, the filing, the decree. The spouses file together as co-petitioners with the agreement attached, which is also the most private route through the courthouse.
The sequence in step six is deliberate. The money is easier once both parents know the children are settled.
Why Mediations Fail
Mediators tend to describe failure in terms of personality: one spouse was inflexible, emotions ran high, the parties could not communicate. Those are real, but in my experience they are usually symptoms. The underlying causes are more concrete, and most of them are preventable.
Negotiating before the numbers exist. A couple with a house, a 401(k), and a business nobody has valued cannot settle intelligently until the third number exists. Mediations that start with positions instead of appraisals stall when the first real figure arrives and one spouse realizes the deal they were discussing was built on a guess.
Incomplete disclosure. This is the most common fatal problem and the one with the longest tail. In Colorado, a settlement built on a material omission can be reopened for five years after the decree under C.R.C.P. 16.2(e)(10) and C.R.S. § 14-10-122. In In re Marriage of Hunt (2015 COA 58), the husband withheld records bearing on his business's value; the mediated memorandum recited that each side had the chance to do due diligence, and the Court of Appeals held the wife could still seek to set it aside. There is no advantage in hiding anything, and an omission discovered mid-mediation ends the mediation.
A power imbalance nobody addresses. When one spouse controlled the finances during the marriage and still controls the information, the other cannot negotiate freely. Full disclosure and the independent review at the end exist to correct for this. Where the imbalance is more than informational, where there are allegations of domestic violence or abuse, mediation is the wrong process, and a careful mediator screens for it at intake and declines.
Expectations the law will not support. A spouse who expects to keep everything, or to deny the other parent meaningful time, is not negotiating; they are waiting for the other side to surrender. The cure is a mediator willing to say, in front of both of them, that a court would never order it. That candor is uncomfortable, and it is the point.
Treating the signed memorandum as a draft. Under Yaekle v. Andrews (Colo. 2008), a signed mediated memorandum can bind the parties as a contract before a court ever sees it. People who sign a one-page MOU at nine at night on the theory that the "real" agreement comes later are often surprised to learn the deal is already done. The better practice is to leave mediation with the complete agreement, not a summary of one.
What the Research Actually Shows
Most of what gets written about mediation outcomes is marketing. One study is not, and it is worth knowing about.
Robert Emery and colleagues at the University of Virginia took families who had filed for contested custody hearings and randomly assigned them to mediate or to litigate, then followed them for twelve years. The results were published in the Journal of Consulting and Clinical Psychology in 2001. Compared with the families who litigated, the nonresidential parents who had mediated were more involved in their children's lives twelve years later, saw them more often, and had more say in parenting decisions, and that increased involvement did not come with more conflict between the parents. Thirty percent of nonresidential parents in the mediation group were seeing their children weekly or more; in the litigation group the figure was nine percent. Thirty-nine percent of nonresidential parents who had litigated had seen their children once or less in the prior year, against fifteen percent of those who had mediated.
The mediation in that study averaged about five hours. It did not change the terms of the agreements much; the custody arrangements looked similar in both groups. What it changed was how the parents treated each other afterward. That is the finding worth remembering: the process by which a parenting plan is reached shapes how it is lived for a decade.
Emery's own review of the broader literature puts agreement rates across mediation studies somewhere between forty and eighty percent depending on the population and how success is counted, with satisfaction consistently higher among people who mediated than among people who litigated. Those are the honest numbers. Anyone quoting a single tidy figure is selling something.
The Question to Ask Your Mediator
Mediation literature spends a lot of time on style: facilitative mediators who keep their opinions to themselves and guide the conversation, evaluative mediators who tell the parties how a court would likely rule. The debate is largely academic. In divorce, every couple eventually asks the same question: what would a judge do with this? A mediator who cannot answer it is not neutral; they are uninformed, and an uninformed neutral favors whichever spouse is the better negotiator.
So the question to ask is not "what is your style?" It is "how many of these cases have you tried?" A mediator who has sat in family courtrooms for thirty years and tried seventy-plus cases to a judge knows what the alternative to settlement looks like, because they have lived it. They can tell both spouses, in the same room and the same words, what a Colorado court is likely to do with the parenting plan, the house, the business, and the maintenance claim, and what each proposal is worth after tax. Couples settle when both of them believe the number in front of them is close to what a courtroom would produce at a fraction of the cost. The mediator's job is to make that belief accurate.
A mediator who has never tried a case can only guess at the alternative. In a process where the alternative is the only real leverage either spouse has, that is a serious gap.
Who Mediation Is For
It fits when both spouses want the divorce resolved fairly and would rather decide the terms themselves than have a judge decide them; when both can sit in a room, or on a screen, and talk; when both will disclose everything and expect the same of the other; and when both want the children's arrangements built by the two parents rather than by an evaluator.
It does not fit when there are allegations of domestic violence or abuse, when one spouse will not disclose or has moved assets, when one spouse is using the process to delay or to learn the other's case, or when either spouse wants a lawyer on their side in the room. That last one is not a criticism. It is a different service, and a good mediator will say so and refer the couple to counsel.
Talk to a Denver Divorce Attorney and Mediator
Aaron Herzberg is a divorce and custody trial lawyer in Greenwood Village, Colorado, serving Denver and the Front Range, with more than thirty years of trial experience, a Georgetown LL.M. in Taxation, and two decades of owning and running businesses. He serves as a neutral mediator for couples who approach him together, and represents individual spouses in mediation and litigation when they come to him alone. The two roles never overlap in the same case.
Read the full guide on DenverDivorce.com: Divorce Mediation with Aaron Herzberg as the Neutral, including the rules he mediates under, the fee structure, and the joint request form. If you are reading this alone and want a lawyer on your side, start instead with Mediation & Litigation, as Your Lawyer.
Educational information only, not legal advice. Reading this article does not create an attorney-client relationship.
