(dissenting):
Although I agree that defendant Cay-etano Morin-Degollado’s conviction should be affirmed, I must respectfully dissent from the majority’s affirmance of co-defendant Telesforo Ray Gomez’s conviction. As the majority points out, the district court admitted hearsay testimony of an informant’s statement that the owners of the marijuana would be driving a red 1974 Ford sedan. I fully *422agree with the majority that it was error to admit this testimony because, as Judge Roney notes,
“ ‘[it] pointed directly to the suspects involved’ . . . In any kind of balancing approach, we would hold that the need for the evidence does not outweigh the possible improper prejudice that results from its admission.”
However, the matter is not concluded with this finding. Relying on United States v. Rodriguez, 5 Cir. 1975, 524 F.2d 485, the majority observes that this clearly wrongful admission of hearsay testimony requires reversal only if there is a “significant possibility that . the statement had a ‘substantial impact’ on the verdict of the jury.” The relevant portion of the cited Rodriguez opinion is as follows:
“To require a new trial, the prejudicial effect of improper matter, viewed in the context of that particular trial, must not be overwhelmed by evidence of guilt. A significant possibility must exist that, considering the other evidence presented by both the prosecution and the defense, the statement had a substantial impact upon the verdict of the jury.”
Where I differ from the majority is in the application of the Rodriguez standard. In effect, Judge Roney concludes that the evidence against defendant Gomez is of such overwhelming strength that there could be no “significant possibility” that the erroneously admitted testimony could have had a “substantial impact” on the jury verdict. I believe that this conclusion is supportable as to defendant Cayetano Morin-Degollado who owned and drove the red Ford. The evidence against the back seat passenger, defendant Gomez, however, is, as the majority twice admits, “not as strong.” That evidence, although perhaps sufficient to support a guilty verdict,1 is in fact very weak, and certainly does not “overwhelm” the prejudicial effect of the improper matter. See Rodriguez, supra, at 487. Absent the informant’s testimony, the evidence shows only that Gomez was in the company of persons dealing in marijuana. Nowhere does the record indicate that Gomez left the bar to accompany Howison and Josué Morin-Degollado when they went into the woods to load the hidden marijuana into a pickup truck. The Ford sedan in which Gomez was a passenger was not the vehicle in which the police found the illegal weed. The blanket in Cayetano Morin-Degollado’s car was never connected in any way with Gomez nor was the gun. We have no testimony of Gomez’s participation in any marijuana dealings. The case against Gomez rests entirely on the fact that he kept company with individuals involved in illegal drug transactions — none of which transactions were shown to have occurred in his presence.
The majority attempts to minimize the impact of the hearsay testimony by noting first that “[ojwnership of the marijuana was not an issue.” Whatever the legal distinctions between ownership and possession, it seems indisputable to me that since the Government’s theory is one of constructive possession, there must be a significant possibility that the hearsay as to ownership had a substantial impact on the jury — particularly in a *423prosecution as close as this one. Moreover, the majority’s contention that “the [hearsay] statement [with respect to ownership] was not directly controlling on any point in issue” and that therefore there is no significant possibility that the evidence had a substantial impact on the jury is strangely at odds with the majority’s earlier assertions where it is said “That testimony tended to point directly to a connection between the marijuana and the occupants of the red Ford,” and “ ‘[the testimony] pointed directly to the suspects involved.’ ”
The majority’s second effort to buttress its conclusion is found in its statement “Direct eyewitness testimony overwhelmingly tied those in the Ford to the marijuana.” As I understand the evidence, no eyewitness testified that Gomez had been seen in the presence of the marijuana or that he had participated in any transactions. Rather the testimony established that Gomez was in the back seat of the car that picked up Josué Morin-Degollado who had handled the illegal goods. In my opinion, this evidence does not “overwhelmingly” tie Gomez to the marijuana dealings, and thus does meet the Rodriguez standard for reversal.2
A comparison with Rodriguez where Judge Wisdom found the other evidence of guilt so convincing that the erroneous submission could not have any “substantial impact,” is instructive. In Rodriguez, an agent testified to an informant’s statement that the defendant was the owner of the marijuana. Then, a convicted co-conspirator testified
to the details of the operation, implicating Rodriguez in numerous ways. Not only did he testify that he and Rodriguez were the owners of the marijuana, but he gave details of its purchase, importation, and eventual delivery to the appellant’s vehicle where it was later seized.
Rodriguez, supra, at 486. Finally, the evidence showed that Rodriguez was the owner of the vehicle in which the police seized the marijuana. In contrast, here no one testified that Gomez was involved in the transactions. The red Ford sedan in which Gomez rode was not the truck from which the police confiscated the 460 pounds of marijuana, nor did Gomez own or drive the Ford.
Because this case requires a completely factual analysis for its result, my dissent is in one sense reluctantly voiced. In another sense, however, I am compelled to speak out for behind this evi-dentiary screen, we encounter what I perceive as a dangerous first step in the direction of debilitation and eventual erosion of the rule against hearsay testimony. “Substantial impact” and “overwhelming evidence” are concepts often difficult in application. But we must not permit their use on a case by case basis to undermine our long revered tradition against out-of-court statements lacking the necessary special indicia of trustworthiness. Clearly the evidence here is not overwhelming. In fact, it is barely sufficient to sustain these convictions. The majority’s finding of no “substantial impact” in the face of such weak evidence has the unfortunate effect of importing wholesale the broad sufficiency test of Glasser v. United States, 315 U.S. 60, 80, 62 S.Ct. 457, 469, 86 L.Ed. 680, 704 (1941), into an area which I believe was intended to be much *424more narrowly circumscribed. See note 1 supra. Certainly Rodriguez did not perceive the Glasser formulation to be a curative incantation in the hearsay domain. Therefore, let us not by wrongfully equating these two standards travel the road to conviction while remaining insensitive to the legal rules governing our journey.
Finally, in considering the claims raised here, we should not forget that the real problem in Rodriguez and the present case stems from trial court’s originally having permitted the introduction of incurable hearsay testimony. Trial courts must carefully guard against the initial admission of free-wheeling and clearly unfair hearsay admissions of informant’s statements. They must be quick to abjure this testimony which so often endangers a defendant’s right to a fair trial with full rights of confrontation. Moreover, by adherence to this adjuration, burdensome and often difficult ad-hoc appellate review of individual cases will become a judicial rarity.
I respectfully dissent.
. I cannot conclude in the light of the standard set forth in Glasser v. United States, 315 U.S. 60, 80, 62 S.Ct. 457, 469, 86 L.Ed. 680, 704 (1941), that the evidence here was not sufficient to support a jury verdict of guilty. However, Rodriguez does not utilize the Glasser test. Under Rodriguez, we must reverse the trial court if there is a “ ‘significant possibility . considering the other evidence . [that] the . . . statement had a substantial impact upon the verdict . . ” Thus, a reversal may be in order upon a finding of a significant possibility of the requisite impact even though the main evidence standing alone could support a conviction. See, e. g., United States v. Ratner, 5 Cir. 1972, 464 F.2d 169, and Odom v. United States, 5 Cir. 1967, 377 F.2d 853. It is in borderline cases, like the present one, exhibiting weak but sufficient evidence, that Rodriguez has its real importance.
. Later in his opinion, Judge Roney outlines additional evidence against Gomez. He points out that “[t]he quantity of contraband involved, 460 pounds, required the assistance of others to distribute” and also that “[t]he area of the bar was a well known smuggler’s point.” I do not believe that either of these assertions adds to the state’s case. The weight of the illegal substance is not evidence of who was involved in the transaction and in this case does not lead to a rational inference that there must have been five men involved as opposed to four or three. In fact, the agent’s testimony indicated that Howison and Josué Morin-Degollado loaded the marijuana into the truck without help from the other defendants. The two of them could and did easily handle the marijuana filled burlap bags without Gomez’s assistance. Moreover, the fact that the area was known as a smuggler’s point does implicate any and all who might be in that area.