Defendant Joshua R. Kilby (“Kilby”) was convicted of conspiracy to distribute and possess with intent to distribute controlled substances in violation of 21 U.S.C. §§ 841 and 846. On appeal, Kilby argues that the evidence was insufficient to sustain his conviction. Kilby also appeals his sentence on three grounds, arguing that the district court erred in: (1) calculating the amount of 5-MEO-DIPT (“Foxy”) attributable to Kilby; (2) deciding the amount of MDMA (“Ecstasy”) attributable to Kilby; and (3) imposing an unreasonable sentence. We have jurisdiction pursuant to 18 U.S.C. § 3742 and 28 U.S.C. § 1291. We affirm Kilby’s conviction, but vacate his sentence and remand for resen-tencing.

BACKGROUND

Prior to 2002, Kilby and Eric Lucker (“Lucker”) worked together to sell Ecstasy. In 2002, Kilby was convicted of drug trafficking in Washington, and was sentenced to two years in prison, which he began serving in October 2002. Kilby and Lucker started to work together again in the summer of 2002, before Kilby started his prison term. Kilby introduced Lucker to several other people, as part of a plan to[*1138] manufacture and sell Ecstasy and Foxy. When Kilby went to prison, Lucker agreed to sell drugs for Kilby and send him payments while he was in jail. Lucker did in fact make several payments to Kilby during 2003, which totaled over $1000.

While in prison, Kilby became friends with John Anthony (“Anthony”), another convicted drug trafficker. Shortly before Anthony’s release in May 2003, Kilby propositioned Anthony about resuming his drug trade in concert with Kilby’s associates. Specifically, Kilby told Anthony to contact Lucker, who would sell Ecstasy to Anthony to get him started again in the business. Anthony played along with Kil-by, but had no intention of going back into the drug trade.

After being released from prison, Anthony contacted Drug Enforcement Administration (“DEA”) Agent Niles Gooding and informed him about his conversations with Kilby. On Agent Gooding’s instructions, Anthony phoned Lucker and inquired about buying Ecstasy. Lucker only agreed to sell Ecstasy to Anthony after obtaining confirmation from Kilby about him. Pursuant to Agent Gooding’s instructions, Anthony told Lucker that he could not travel to Seattle to purchase the drugs, and offered to send his “friend” instead. Anthony’s “friend” was Agent Gooding.

Over the course of several months, Agent Gooding made three purchases from Lucker totaling 2500 pills of Ecstasy. During those purchases, Lucker discussed sending money to Kilby as part of the overall deal. Records show that Lucker sent Kilby money within days of Agent Gooding’s first two purchases. After setting up one more purchase with Lucker, Agent Gooding arrested him in March 2004.

PROCEDURAL HISTORY

Kilby was indicted for conspiracy to distribute and possession with intent to distribute Ecstasy and Foxy in violation of 21 U.S.C. §§ 841 and 846. During the three-day jury trial, after the government rested, Kilby made a motion for acquittal. Kilby argued that the alleged conspiracy had not been proved because Lucker had withdrawn from the conspiracy. The district court denied the motion, stating that the question of withdrawal was a fact to be found by the jury. The jury found Kilby guilty of the conspiracy and determined that he had conspired to distribute/possess with intent to distribute 2500 tablets of Ecstasy and 7000 tablets of Foxy. No Foxy tablets were seized in the case, however, so it was not possible to measure the weight of those tablets.

In the first Presentence Investigation Report (“PSR”), the probation officer estimated the weight of the Foxy tablets by considering the DEA’s recommended dosage of the drug and Foxy’s heightened potency as compared to another known drug, DMT. This calculation, when combined with the amount of Ecstasy, resulted in a base offense level of 24 under the Sentencing Guidelines.

Kilby objected to the first PSR, arguing that it was legally improper to consider the potency of Foxy in determining the weight of the tablets. The probation officer agreed and issued a second PSR. In the second PSR, the probation officer estimated the weight of the Foxy tablets by considering two cases in which actual Foxy tablets had been recovered. In one case, from Virginia, the tablets each weighed 140 milligrams, and in the second case, from California, the tablets each weighed 235 milligrams. Considering these weights together, the officer conservatively estimated the weight of the tablets in this ease to be 100 milligrams each. This cal[*1139] culation resulted in a base offense level of 26.

Kilby objected to the second PSR, arguing that there was an insufficient factual basis for the finding of the weight of the Foxy tablets. The district court nonetheless adopted the calculations in the second PSR, finding that 100 milligrams per Foxy tablet was a conservative estimate that gave the benefit of the doubt to Kilby. The court, however, did not want to punish Kilby for identifying the error in the first PSR. The court also wanted to be very conservative in its estimate of the amount of Foxy. Therefore, the court used a base offense level of 24, as recommended by the first PSR.

Factoring in adjustments for Kilby’s role in the offense and for his criminal history, the court found the applicable Guidelines range to be 92 to 115 months. In imposing the sentence, the district court considered the Guidelines, the policies of the Sentencing Commission, and the factors listed in 18 U.S.C. § 3553(a). The court acknowledged that the Guidelines were advisory, but concluded that the Guidelines range was “appropriate and should be applied in this case.” The court thus imposed a sentence of 96 months’ imprisonment, followed by five years of supervised release. Kilby timely appeals.

ANALYSIS

I. Sufficiency of the Evidence

Kilby contends that there was insufficient evidence to prove the charged conspiracy because Lucker had withdrawn from the conspiracy prior to selling Ecstasy to Agent Gooding. “Where a defendant moves for acquittal at the close of the government’s evidence, we review de novo whether sufficient evidence exists to support a guilty verdict.” United States v. Stewart, 420 F.3d 1007, 1014 (9th Cir.2005). “Evidence is sufficient if, viewed in a light most favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt.” United States v. Odom, 329 F.3d 1032, 1034 (9th Cir.2003).

“A conspirator can withdraw from a conspiracy by: (1) disavowing the unlawful goal of the conspiracy; (2) affirmatively acting to defeat the purpose of the conspiracy; or (3) taking definite, decisive, and positive steps to disassociate himself from the conspiracy.” United States v. Fox, 189 F.3d 1115, 1118 (9th Cir.1999) (internal quotation marks omitted). Kilby bases his argument on Lucker’s testimony that he told Kilby in mid-2003 that he was “fed up with the drug business” and that he was “done dealing drugs.” A review of the record, however, demonstrates that the jury had ample evidence on which to base a finding that Lucker did not in fact withdraw from the conspiracy.

First, Lucker’s own testimony is equivocal. Immediately after he made the statements cited by Kilby, Lucker also testified on cross-examination that he never withdrew from the conspiracy:

Q: And the deal was done, is that correct, because Josh [Kilby] couldn’t sell drugs?

A: The deal never ended. I didn’t talk to Josh about ending the deal.

In addition, Lucker also testified on cross that he talked about “getting out of the drug business” before Anthony contacted him. Lucker therefore confirmed Anthony’s request to purchase Ecstasy with Kil-by after his comments about quitting the drug business. Finally, on redirect examination, Lucker asserted that he had not testified that he told Kilby he “quit the drug business,” and did not recall having a conversation where he told Kilby that he quit. Thus, Lucker himself confirmed that he did not withdraw from the conspiracy, but rather continued his participation in it.

Second, evidence of Lucker’s conduct supports the finding that he never with[*1140] drew from the conspiracy. Lucker sent Kilby money both prior to and after his alleged withdrawal. Lucker agreed to sell Ecstasy to Anthony after getting confirmation from Kilby. And Lucker, while selling Ecstasy to Agent Gooding, discussed sending some of the proceeds to Kilby. This evidence, when combined with Luck-er’s self-contradictory testimony, and considered in the light most favorable to the government, is sufficient to support the finding that Lucker did not withdraw from the conspiracy. See Stewart, 420 F.3d at 1015 (“We must respect the province of the jury to ascertain the credibility of witnesses, resolve evidentiary conflicts, and draw reasonable inferences from proven facts, by assuming that the jury resolved all such matters in a manner which supports the verdict.” (internal quotation marks omitted)). Accordingly, we affirm Kilby’s conviction.

II. Review of Kilby’s Sentence

A. Standard and Method of Review

When we review a sentence, the first step is to determine if the district court made a material error in its Guidelines calculation that serves as the starting point for its sentencing decision. United States v. Cantrell, 433 F.3d 1269, 1280 (9th Cir.2006). If there was material error in the Guidelines calculation, we will remand for resentencing, without reaching the question of whether the sentence as a whole is reasonable. Id. If the district court committed no error in applying the Guidelines, we will next consider challenges to the reasonableness of the overall sentence in light of the factors specified in 18 U.S.C. § 3553(a). Id.

With regard to the Guidelines calculation, we review de novo whether the district court applied the correct burden of proof in attributing drug quantity to a defendant. United States v. Banuelos, 322 F.3d 700, 704 (9th Cir.2003). We review a district court’s interpretation of the Guidelines de novo. Cantrell, 433 F.3d at 1279. We also review de novo whether a district court’s method of approximation of the relevant drug quantity conforms to the Guidelines. United States v. Rosacker, 314 F.3d 422, 425 (9th Cir.2002). The district court’s factual findings are reviewed for clear error. Cantrell, 433 F.3d at 1279.

B. Determination of Amount of Foxy

Kilby contends that the district court erred in estimating the amount of Foxy attributable to him in its Guidelines calculation because the method of estimation was unreasonable and did not amount to proof beyond a reasonable doubt. As an initial matter, Kilby’s contention that the drug quantity finding must have been proved beyond a reasonable doubt is mistaken. We have held that following United States v. Booker, 543 U.S. 220, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005), district courts should resolve factual disputes at sentencing by applying the preponderance of the evidence standard. See United States v. Ameline, 409 F.3d 1073, 1086 (9th Cir.2005) (en banc) (citing United States v. Howard, 894 F.2d 1085 (9th Cir.1990)); see also United States v. Dare, 425 F.3d 634, 642 (9th Cir.2005) (“As a general rule, the preponderance of the evidence standard is the appropriate standard for factual findings used for sentencing.”). ^[1] Therefore, the district court did not err by failing to[*1141] require that the quantity of Foxy be proved beyond a reasonable doubt.

Kilby argues that, even if’ the correct standard is preponderance of the evidence, the district court’s finding ’ did not meet that standard. Specifically, Kilby contends that the district court’s method of approximating the weight of the Foxy tablets was not sufficiently reliable.

The applicable Guidelines range for a drug trafficking crime depends upon the weight of drugs involved. United States v. Culps, 300 F.3d 1069, 1076 (9th Cir.2002); U.S.S.G. § 2D1.1(a), (c) (2004). With tablets or pills, the entire weight of the tablet is used in the Guidelines calculation, not just the weight of the pure controlled substance. United States v. Crowell, 9 F.3d 1452, 1454 (9th Cir.1993). Where none of the- drugs has been seized, the district court may approximate the weight of the drugs. See United States v. August, 86 F.3d 151, 154 (9th Cir.1996); U.S.S.G. § 2D1.1, cmt. n. 12. ^[2]

“Approximations of drug quantity must meet three criteria.” Culps, 300 F.3d at 1076. First, “the government is required to prove the approximate quantity by a preponderance of the evidence ... [which means that][t]he district court must conclude that the defendant is more likely than not actually responsible for a quantity greater than or equal to the quantity for which the- defendant is being held responsible.”. Id. (internal quotation marks and citation omitted). Second, “the information which supports an approximation must possess sufficient indicia of reliability to support its probable accuracy.” Id. (internal quotation marks omitted). Third, since the “sentence depends in large part upon the amount of drugs .,. and approximation is by definition imprecise, the district court must err on the side of caution” in approximating the drug quantity. Id. 3

We have not previously evaluated an approximation based upon the weights of drugs seized in completely unrelated cases. We have approved several different methods for making drug quantity approximations, but in these methods, the variables involved were all based on facts specific to the defendant’s case. See, e.g., United States v. Alvarez, 358 F.3d 1194, 1213 (9th Cir.2004) (approving approximation based on co-conspirator testimony); Culps, 300 F.3d at 1077 (approving approximation based on multiplying an estimated daily or weekly quantity by an applicable period of time); August, 86 F.3d at 154-55 (approving approximation based on lab capacity); United States v. Basinger, 60 F.3d 1400, 1409-10 (9th Cir.1995) (approving approximation based on two empty, one-pound containers of ephedrine); see also United States v. Gonzalez-Sanchez, 953 F.2d 1184, 1187 (9th Cir.1992) (citing cases that approve of the method of converting cash proceeds into equivalent amount of drugs);[*1142] U.S.S.G. § 2D1.1, cmt. n. 12 (stating that, when approximating quantity, a court may consider “similar transactions in controlled substances by the defendant” (emphasis added)).

In contrast, the district court in this case based its approximation on the weights of Foxy tablets seized in two unrelated cases from different parts of the country. Based on these circumstances, there is no reliable way to know whether the Foxy tablets in this case had a similar weight to the tablets in the other cases. The government presented no evidence that Kilby’s case was in any way related to these other Foxy cases, such as evidence of a common supplier. Nor did it present any evidence that Foxy tablets are always the same approximate size. To the contrary, the statistical “sample” only had two pieces of data, and the weights in those cases were quite different. Cf. United States v. Scheele, 231 F.3d 492, 500 (9th Cir.2000) (“[T]he district court’s failure to consider the margin of error when arriving at the quantity of drugs on which the sentence was based constitutes error .... ”). Therefore, we conclude that the district court’s method of approximation did not possess sufficient indicia of reliability to serve as a basis for calculating the quantity of Foxy. See Culps, 300 F.3d at 1076. ^[4] Accordingly, we hold that the district court clearly erred in its finding of the amount of Foxy involved.

C. Determination of Amount of Ecstasy

Kilby contends that the district court erred in attributing 2000 Ecstasy pills to him for sentencing purposes. Specifically, he argues that such a quantity was not within the scope of any agreement between him and Lucker, and that Luck-er’s sales were not reasonably foreseeable to Kilby.

“[I]n determining for purposes of sentencing the quantity of drugs for which a conspirator will be held responsible, the district court is required to determine the quantity of drugs the conspirator reasonably foresaw or which fell within the scope of his particular agreement with the conspirators.” Banuelos, 322 F.3d at 702 (internal quotation marks omitted). The record amply demonstrates that Lucker’s sales of Ecstasy were reasonably foreseeable to Kilby. Kilby specifically counseled and induced Anthony to buy “two boats” (2000 pills) of Ecstasy from Lucker. In addition, Lucker only agreed to sell Ecstasy to Anthony after obtaining confirmation from Kilby. Therefore, Lucker’s Ecstasy sales were specifically contemplated by Kilby. Accordingly, the district court did not err in attributing the 2000 Ecstasy pills to Kilby for sentencing purposes.

CONCLUSION

The district court clearly erred in its Guidelines calculation by using an insufficiently reliable method of approximating the amount of Foxy attributable to Kilby. The district court’s error was not harmless because, had it found the tablets to each weigh 10 milligrams instead of 100 milligrams, Kilby’s base offense level would have been 20 instead of 24, the level assigned by the district court. See U.S.S.G.[*1143-1153] § 2Dl.l(c). We therefore vacate the sentence and remand for resentencing without reaching the question of whether the sentence as a whole was reasonable. See Cantrell, 433 F.3d at 1280.

Conviction AFFIRMED; sentence VACATED and REMANDED for resentencing.