Shawn Quinton Regan drove vehicles transporting cocaine cross-country and handled proceeds from drug sales in an expansive drug conspiracy involving kingpin Ralph Duke. See United States v. Duke, 940 F.2d 1113 (8th Cir.1991). At Regan’s trial with Duke and four others, a jury convicted Regan of aiding and abetting the distribution of cocaine, distribution of cocaine, and conspiracy to possess with intent to distribute cocaine. See 21 U.S.C. §§ 841(a)(1), 846 (1988). The district court sentenced Regan to five concurrent terms of 235 months imprisonment. Regan appeals his convictions and sentences. We affirm.

First, Regan asserts there was a variance between the single conspiracy charged in the indictment, and the proof at trial of at least two independent conspiracies: the “reverse buy conspiracy” and the “Columbian distribution conspiracy.” Re-gan did not participate in the reverse buy transaction. He argues the variance affected his substantial rights because the “spillover” of the reverse buy evidence at his trial prejudiced him. See United States v. Jones, 880 F.2d 55, 66 (8th Cir.1989). We disagree.

The question whether a single conspiracy or multiple conspiracies exist is for the jury. United States v. Zimmerman, 832 F.2d 454, 457 (8th Cir.1987) (per curiam). If the record contains evidence from which the jury could find one overall agreement to commit an illegal act, the evidence establishes a single conspiracy. United States v. Baker, 855 F.2d 1353, 1356 (8th Cir.1988), cert. denied, 490 U.S. 1069, 109 S.Ct. 2072, 104 L.Ed.2d 636 (1989). The existence of a single agreement can be inferred when the evidence reveals the participants shared a common aim or purpose and mutual dependence and assistance existed. United States v. Lee, 782 F.2d 133, 134 (8th Cir.1986). Considering the evidence in the light most favorable to the verdict, we conclude a reasonable jury could have determined Regan participated in a single, extended conspiracy to distribute cocaine in the Twin Cities area. Both asserted conspiracies involved common participants and a common goal. Thus, Regan’s nonparticipation in one of the many transactions undertaken as part of the single conspiracy’s common plan does not demonstrate there was a variance between the indictment and the proof offered at trial. See id. at 135.

Second, Regan argues the evidence was insufficient to support the jury’s verdict on the distribution counts. Viewing the evidence in the light most favorable to the verdict and giving the government the benefit of all reasonable inferences that logically may be drawn from the evidence, we conclude the evidence against Regan on these counts amounted to more than mere presence at the scene. See United States v. Matlock, 786 F.2d 357, 360-61 (8th Cir.1986). A witness’s testimony suggested Regan delivered the cocaine for one of the transactions. Thus, we conclude the evidence was sufficient to support Regan’s distribution convictions.

Third, Regan asserts the trial court committed error in giving the jury a willful[*1136] blindness instruction. This instruction is particularly appropriate when the defendant denies any knowledge of a criminal scheme despite strong evidence to the contrary. United States v. Massa, 740 F.2d 629, 643 (8th Cir.1984), cert. denied, 471 U.S. 1115, 105 S.Ct. 2357, 86 L.Ed.2d 258 (1985). The instruction should not be given, however, when the evidence “points solely to either actual knowledge or no knowledge of the facts in question.” United States v. Hiland, 909 F.2d 1114, 1130 (8th Cir.1990).

Here, Regan admitted involvement with the coconspirators and acts of the conspiracy, but denied any knowledge the acts concerned cocaine. Nevertheless, substantial evidence indicated Regan knew he was dealing with drugs. In light of the instructions as a whole, we conclude the district court did not commit error in giving the willful blindness instruction. Id. at 1131 (willful blindness instruction properly given when evidence supports inference of deliberate ignorance). Further, any error in giving this instruction was harmless in light of the overwhelming evidence of Regan’s involvement.

Fourth, Regan contests his base offense level arguing the government did not prove by a preponderance of evidence that he was involved in the distribution of seventy-seven kilograms of cocaine. The trial testimony of three witnesses, however, attributed Regan with handling at least fifty kilograms of cocaine. Because a base offense level of thirty-six applies to amounts of at least fifty kilograms but less than one hundred fifty kilograms of cocaine, U.S. S.G. § 2D1.1(c)(4), we need not resolve whether the government proved the difference between fifty and seventy-seven kilograms. Thus, we conclude the district court properly assigned Regan a base offense level of thirty-six.

Finally, asserting he was merely a courier, Regan challenges the district court’s failure to decrease his offense level by two as a minor participant. See U.S. S.G. § 3B1.2. We review a sentencing court’s denial of this reduction for clear error. United States v. Olson, 931 F.2d 1250, 1252 (8th Cir.1991). Having examined the evidence, we conclude Regan was more than a mere courier. Indeed, Regan was deeply involved in the transportation and handling of a large quantity of drugs and cash with his coconspirators. In addition, Regan received economic benefits from his involvement in the conspiracy— Ralph Duke paid Regan’s rent in an expensive high-rise apartment. Thus, the district court committed no error in denying Regan a reduction in his offense level as a minor participant.

Accordingly, we affirm Regan’s convictions and sentences.