Kim Darby Saenz pled guilty to one count of conspiracy to distribute marijuana. At sentencing, the government moved, pursuant to USSG § 5K1.1 and 18 U.S.C. § 3553(e), to reduce her sentence below the applicable sentencing guideline range of 63 to 78 months and the statutory minimum sentence of 60 months. The district court granted the motion and sentenced Saenz to 20 months’ imprisonment. The government appeals the extent of the reduction in sentence, and we reverse.

I.

Pursuant to a plea agreement with the government, Saenz pled guilty to one count of conspiracy to distribute marijuana. Saenz was charged after officers discovered marijuana in two vans located in a parking lot of a hotel in Onawa, Iowa. Saenz and her husband, Rudolph, were staying in a hotel room directly in front of the vans, and officers confronted them after finding the marijuana. After the Saenzes admitted to traveling in one of the vans, they were placed under arrest. A man and woman, Veronica Rodriguez-Cortez and Jose Rodriguez-Medrano, were staying in the room next to the Saenzes, and they were also arrested. When the[*1161] foursome was interviewed, law enforcement officers learned that a third vehicle was traveling with the group. Officers located this vehicle, and a passenger, Christian Jimenez of San Diego, eventually was charged together with Rodriguez-Me-drano.

Rodriguez-Cortez later testified before the grand jury that Rodriguez-Medrano had approached her and asked that she transport two vans filled with marijuana from Chula Vista, California, to Sioux City, Iowa. Rodriguez-Cortez said that she told Saenz about the trip, and Saenz and her husband drove the second van. Rodriguez-Medrano urged them to bring their children, in order to disguise the purpose of their travel, and agreed to pay each of them $2500 for making the trip. Rodriguez-Medrano, Rodriguez-Cortez, Rudolph Saenz, and Jimenez eventually all pled guilty to drug trafficking offenses.

A presentence investigation report recommended a sentencing range of 63 to 78 months’ imprisonment for Ms. Saenz under the then-mandatory United States Sentencing Guidelines. At sentencing, the government moved to reduce the sentence under USSG § 5K1.1 and 18 U.S.C. § 3553(e), based on Saenz’s provision of substantial assistance, and recommended a departure of 30 percent, or 19 months, to a final sentence of 44 months’ imprisonment.

In support of its substantial-assistance motion, the government described Saenz’s cooperation. According to the government, Saenz talked to law enforcement officers about her co-conspirators on the same day that she was arrested. Within a day or two, the information provided by Saenz was used, along with statements obtained from the co-conspirators, to assist in the preparation of affidavits in support of criminal complaints. Saenz later testified at the sentencing hearing for Rodriguez-Medrano, where she corroborated testimony of Rodriguez-Cortez regarding the use of minors to avoid detection. Saenz also attempted to cooperate with Drug Enforcement Administration agents in San Diego while she was on pre-trial release, and assisted with the procurement of one search warrant, but the information that she provided was outdated and did not lead to any arrests or seizures.

The court granted the government’s substantial-assistance motions. In the court’s view, Saenz was “exceptionally timely” in her cooperation and there was “no indication that she was anything but totally truthful, complete, and reliable and that she gave them all the information she could.” (S. Tr. at 16). The court also opined that the government’s percentage recommendations for departures were “arbitrary and capricious and without any basis because they fail to disclose how they arrive at their decision.” (S. Tr. at 17). The court expressed its view that “any defendant who is timely, completely truthful, complete, reliable, and tells the government everything they need to know deserves more than 50 percent” reduction. (I'd). Citing its evaluation of the factors set forth in USSG § 5K1.1, the court reduced Saenz’s sentence to 20 months’ imprisonment.

II.

We recently reviewed three cases involving the reductions of sentences based on the provision of substantial assistance, and we set forth certain parameters regarding appellate review of such sentences. See United States v. Dalton, 404 F.3d 1029 (8th Cir.2005); United States v. Haack, 403 F.3d 997 (8th Cir.2005); United States v. Pizano, 403 F.3d 991 (8th[*1162] Cir.2005). ^[1] These decisions explain that the court of appeals is charged with considering whether the extent of a reduction in the now-advisory guideline sentence or below the statutory minimum sentence is “reasonable,” and that we review the district court’s decision for abuse of discretion. E.g., Dalton, 404 F.3d at 1032.

We said in Dalton that “[a]n extraordinary reduction must be supported by extraordinary circumstances,” id. at 1033, and we find that maxim applicable here. Departures under § 5K1.1 and reductions under § 3553(e) should not be untethered from the structure of the advisory guidelines. They take place, rather, within the framework of an advisory guideline scheme designed to reduce unwarranted sentence disparities among similar defendants, see 18 U.S.C. 3553(a)(6); United States v. Booker, 543 U.S. 220, 125 S.Ct. 738, 767, 160 L.Ed.2d 621 (2005), and we are mindful that the Sentencing Commission has concluded that most adjustments for aggravating or mitigating circumstances should be in the amount of two, three, or four offense levels. Cf. United States v. Ferra, 900 F.2d 1057, 1064 (7th Cir.1990) (“The Commission prescribes two-level adjustments for relatively serious offense characteristics, and departures of more than two levels should be explained with a care commensurate with their exceptional quality.”). In view of this context, we indicated in Dalton that a 75 percent (or 12-level) downward departure was “extraordinary,” 404 F.3d at 1033, and cited favorably our observation elsewhere that a 50 percent downward departure was an “extraordinary sentence reduction.” United States v. Enriquez, 205 F.3d 345, 348 (8th Cir.2000). The reduction granted to Saenz is in the same category. It amounted to the equivalent of an 11-level departure under the advisory guidelines, and a 68 percent reduction in sentence from the otherwise applicable advisory range.

The extent of a departure or reduction pursuant to § 5K1.1 or § 3553(e) “can be based only on assistance-related considerations,” United States v. Pepper, 412 F.3d 995, 998 (8th Cir.2005), and our review of a reduction for substantial assistance typically centers on the non-exhaustive list of factors set forth in § 5K1.1, which the district court should consider in making its determination:

(1) the court’s evaluation of the significance and usefulness of the defendant’s assistance, taking into consideration the government’s evaluation of the assistance rendered;

(2) the truthfulness, completeness, and reliability of any information or testimony provided by the defendant;

(3) the nature and extent of the defendant’s assistance;

(4) any injury suffered, or any danger or risk of injury to the defendant or his family resulting from his assistance;

(5) the timeliness of the defendant’s assistance.

USSG § 5Kl.l(a).

We are troubled in this case by the district court’s statement that “any defendant who is timely, completely truthful, complete, reliable, and tells the government everything they need to know deserves more than 50 percent” in reduction of sentence from the applicable guideline range.' Timeliness and truthfulness are[*1163] indeed two of the relevant factors for consideration, but we respectfully disagree with the district court’s apparent view that a strong showing in those areas makes reasonable an extraordinary departure or reduction of more than 50 percent, or the equivalent of more than about eight offense levels for this defendant, without regard to the nature and extent of the defendant’s assistance, the significance and usefulness of the assistance, or any danger or risk of injury suffered by the cooperating defendant. As we said in Haack, “[a] departure of this extent leaves little room for greater departures for defendants who actually participate in controlled buys, wear wires, give grand jury and trial testimony, or are subjected to significant risk of injury or death to themselves or their family.” 403 F.3d at 1005-06. Conversely, some cooperating defendants present a close question as to whether their assistance merits a substantial-assistance motion at all, and if United States Attorneys know that timely and truthful cooperation automatically justifies cutting a sentence in half when a motion is filed, then it is reasonable to expect that many of these borderline defendants will fail to qualify when the government makes a “rational assessment of the cost and benefit that would flow from moving.” Wade v. United States, 504 U.S. 181, 187, 112 S.Ct. 1840, 118 L.Ed.2d 524 (1992).

A review of the record in this case shows that while Saenz’s assistance is reasonably viewed as “substantial,” the nature and extent of her assistance was relatively limited, the significance and usefulness of her assistance is relatively modest, and she suffered no apparent danger or risk of injury. Saenz was one of four persons apprehended at the Onawa motel whose interview led to the arrest of Christian Jimenez. She testified as a corroborating witness at the sentencing hearing of Rodriguez-Medrano and provided evidence that supported a two-level adjustment in his guideline offense level. See USSG § 3B1.4. She tried unsuccessfully to assist with an investigation of drug trafficking in California. She did not, however, play a lead role in building a case on another offender, participate in undercover work (such as by wearing a recording device or making controlled purchases), give testimony in a grand jury or at a trial, or experience significant risk of injury or death. See Haack, 403 F.3d at 1005-06; Dalton, 404 F.3d at 1033; cf. Pizano, 403 F.3d at 995-96 (affirming as reasonable a 12-level reduction where defendant provided timely and truthful cooperation, was “key witness” against two co-conspirators, gave testimony that could be instrumental in seizing assets from money laundering scheme, provided debriefings and grand jury testimony regarding both a close family member and a “major figure” in the conspiracy, and put himself and his family at risk of harm from “dangerous people” when he testified).

We recognize that sometimes a defendant’s early cooperation will be so effective that another person will feel compelled to plead guilty on that basis, thus obviating the need for the cooperating defendant to provide testimony or perform other acts of assistance in that matter, and a district court may take account of such circumstances in evaluating the factors under § 5K1.1. Here, however, Saenz was not a key witness who helped bring to justice a major criminal figure. She was a corroborating witness and one of four conspirators who helped to identify a fifth. Of course, as the district court observed, “[t]he fact that she doesn’t know the whereabouts of Osama bin Laden is not her fault,” but the proper analysis under § 5Kl.l(a)(l) focuses on the actual significance and usefulness of the assistance, regardless of the defendant’s desire, effort,[*1164] or “fault.” United States v. Davila, 964 F.2d 778, 786 (8th Cir.1992). A defendant who is eager to cooperate; but in a position to provide only assistance of no use or modest value, will be limited either to a reduction under the safety-valve guideline (if her assistance is not deemed to be “substantial”), see USSG §§ 5C1.2, 2Dl.l(b)(6), or to a substantial-assistance reduction that must be calibrated to account for the limited significance and usefulness of the assistance.

The government stresses, in asserting unreasonableness, that the district court’s reduction in this case greatly exceeded that recommended by the United States Attorney, but we are not convinced by that point of argument. It is true, of course, that the district court must give “Substantial weight” to the government’s “evaluation of the extent of the defendant assistance, particularly where the extent and value of the assistance are difficult to ascertain.” USSG § 5K1.1, comment, (n. 3) (emphasis added); Pizano, 403 F.3d at 996. The government, after all, is often in a superior position to know whether a particular cooperating defendant provided the crucial break in an investigation, or the critical testimony that secured either convictions or the cooperation of additional material witnesses, so the command for deference in that regard is quite sensible.

We are less persuaded that the court must give substantial weight to the government’s valuation of the assistance, particularly where the government does not adequately explain its reasoning. Haack, 403 at 1005 n. 2; cf. Pizano, 403 F.3d at 996. While we recognize that a single United States Attorney brings to bear-the broad perspective of one who has evaluated numerous cooperating defendants within a judicial district (indeed, at least during comparable tenure, a broader perspective than a single district judge in a multi-judge district), we have no assurance that United States Attorneys in different districts apply consistent methodologies for valuing substantial assistance and arriving at sentencing recommendations. Judicial deference to wildly varying recommendations in different districts would create, rather than avoid, unwarranted sentence disparities. Even if a consistent practice were shown within the Department of Justice, we would have to consider whether the' degree of reductions recommended under such an approach would be consistent with the structure and theory of the guidelines. If, for example, the United States Attorney in a particular district were to recommend that every defendant who provides timely and truthful cooperation should receive a sentence reduction of greater than 50 percent, we do not believe such a recommendation would be entitled to “substantial weight” any more than we think a district court’s application of such a hard-and-fast rule would be reasonable. Likewise, if a United States Attorney makes identical recommendations of slight departures in cases that are dissimilar, in ways that are material to the appropriate degree of reduction, then neither are those recommendations due much weight from the court. Cf. Haack, 403 F.3d at 1005 n. 2.

Our decision in this case, therefore, turns not on the degree to which the district court’s decision varies from the government’s recommended reduction, but rather on our independent conclusion that the degree of reduction is not reasonable in light of the- evidence concerning the defendant’s assistance, the factors set forth in § 5K1.1, and the overall structure and theory of the guidelines, including the statutory command to reduce unwarranted sentence disparities. The appropriate degree of sentencing reduction cannot be calculated with “mathematical precision,” Haack, 403 F.3d at 1005, and there is a[*1165] range of reasonableness available to the district court in any given case. On this record, however, we conclude that the district court’s analysis was flawed by its conclusion that timely and truthful cooperation always warrants a reduction of more than 50 percent, and that the degree of reduction was excessive and unreasonable under the circumstances of this case. Accordingly, the judgment of the district court is vacated, and the case is remanded for resentencing consistent with this opinion. ^[2]