Defendant appeals from a judgment of conviction entered upon a jury verdict finding him guilty of first degree robbery. (Pen. Code, § 211.)
At the police lineup defendant did not have the assistance of counsel, a right that has since been held to be guaranteed by the Sixth and Fourteenth Amendments to the United States Constitution. (United States v. Wade (1967) 388 U.S. 218 [18 L.Ed.2d 1149, 87 S.Ct. 1926] ; Gilbert v. California (1967) 388 U.S. 263 [18 L.Ed.2d 1178, 87 S.Ct. 1951].) We have heretofore decided, however, that the rule of Wade and Gilbert, requiring the exclusion of in-court identification testimony tainted by displaying an accused to identifying witnesses before trial in the absence of his counsel, unless an effective waiver has been procured, is to be given only prospective application. (People v. Feggans (1967) 67 Cal.2d 444 [62 Cal.Rptr. 419, 432 P.2d 21].) Since this is a preWade and Gilbert case, before defendant may invoke an exclusionary concept he must demonstrate that the lineup “resulted in such unfairness that it infringed his right to due process of law.” (Stovall v. Denno (1967) 388 U.S. 293, 299 [18 L.Ed.2d 1199, 1205, 87 S.Ct. 1967].) Defendant Caruso has made a showing that the lineup was unfair, resulting in tainted in-court identification essential to the People’s case. The People fail to demonstrate that the error was harmless (Chapman v. California (1967) 386 U.S. 18, 24 [17 L.Ed.2d 705, 710, 87 S.Ct. 824]), and the judgment must therefore be reversed.
*185On the morning of February 23, 1965, Olen Seeley, an employee at a Robert Hall store, momentarily left his job to seek some matches in his car. As he walked through the store parking lot he noticed a parked black two-door Ford hardtop with a man inside; he detected nothing unusual about the man or the vehicle.
Later that day, about 12:45 p.m., Seeley and Robert Butkus, the store manager, left the store to make a routine deposit of the morning receipts. Butkus had just inserted the key in the ignition of his station wagon when he heard a car pull up behind him, and both he and Seeley turned to look at it. The car slowed to a halt immediately behind the station wagon and blocked its exit from the store parking lot.
Before the car came to a complete stop the passenger door was opened, and a man wearing a dark zip-up jacket, a snap brim hat, and a handkerchief over the bridge of his nose, quickly approached the driver’s side of the station wagon. As he advanced he drew his hand from the inside of his jacket, disclosing what appeared to be an automatic pistol.
At this point Seeley observed the profile of the driver of the ear, and Butkus “saw” him. Both testified that the length of their view was five to six seconds. Both described the driver as big, with dark wavy hair and a dark complexion. Neither victim could describe the driver’s clothing, although Butkus thought he was wearing a T-shirt. Their attention was primarily focused, quite naturally, on the armed, masked man who was moving toward them. When the masked man reached the station wagon, he ordered Butkus to give him the money and the keys to his car. Butkus quickly complied with both orders, and in response to a further demand he and Seeley crouched down on the seat while the robbers drove off. The total episode, resulting in seizure of a white bag that contained the store’s receipts, had taken no more than 30 seconds. Butkus and Seeley immediately reentered the store and called the police. Officers arrived within minutes, about 12:58 p.m.
Later that afternoon the police located the black Ford approximately four blocks from the scene of the robbery in a parking area reserved for a commercial laboratory. As the laboratory’s production manager, Tom Dulin, returned from lunch he discovered to his annoyance that a car had just parked in his private space. He attempted to explain that it was illegally parked but the driver hurried off, mumbling that *186he would be back later. He noted that the man was carrying a brown paper sack but could not recall the man’s physical appearance. Dulin watched him leave the parking area, run diagonally across Valley Street to a nearby intersection, and then down Victory Boulevard until he passed from sight. He informed the police of the illegally parked ear, and the police ascertained that the vehicle, which was found to have been stolen, was the black Ford used in the robbery. At trial Dulin was unable to remember when he had returned from lunch, and police records apparently did not reflect the time he placed the telephone call.
Officer Ysmael Torres lived near the same laboratory, and on the day in question he left his home in time to report for a 1 p.m. class at the police station. As he proceeded north on Valley Street, very near his home, he watched a man in a brownish jacket emerge from an alley below the parking area for the laboratory, place a white object that resembled a bag underneath his jacket, and enter the passenger side of the car in front of him. The car, with the two men in it, then proceeded to the intersection of Valley Street and Victory Boulevard, made a regular boulevard stop, and then turned down Victory Boulevard for a short distance before turning off. These events attracted his attention for no reason other than that the pedestrian “appeared to be having some conversation” with the driver of the car in front, although the officer could not overhear its contents. He noted the license number of the car and upon learning of the robbery he reported the license number for investigation. The police learned that this number had been assigned to defendant’s family car.
Since the movements of the person observed by Dulin differed substantially from those of the person observed by Torres, it is evident that the two men could not have seen the same individual or the same events. At trial Dulin was uncertain of the features of the person he observed, and Torres also was unable to recall the physical traits of the pedestrian who had initially attracted his attention. On the other hand, Torres described the driver of the car in front of him as “big shouldered” and “dark haired,” but he did not testify that the driver was defendant. At most, Torres’ observations tended to place defendant’s car in the vicinity of the Robert Hall store at the approximate time of the crime.
From the foregoing summary of the testimony, it is manifest that the jury must have found in the victims’ in-court *187identification persuasive evidence that defendant was the driver of the robbery car.
Defendant was arrested in his home on the night of the robbery, and the next morning he was placed in a lineup with four other men. Both victims, Butkus and Seeley, identified defendant as the driver of the robbery ear. Seeley further identified defendant as the man in the car in the Robert Hall parking lot who had briefly attracted his notice earlier that morning when he went to his car to retrieve some matches.
But the uncontradicted testimony of all those who viewed the lineup demonstrates that it was conducted under circumstances which could only have suggested to Butkus and Seeley that defendant was the man to be charged with the offense. Defendant is of imposing stature, being 6 feet 1 inch tall, and weighing 238 pounds. He is of Italian descent, with a very dark complexion, and has dark wavy hair. The two victims, the officer in charge of the investigation, Sergeant Allen, and defendant all testified that the other lineup participants did not physically resemble defendant.1 They were not his size, not one had his dark complexion, and none had dark wavy hair. During the robbery both Seeley and Butkus had noted the driver’s large size and dark complexion, and, if they were to choose anyone in the lineup, defendant was singularly marked for identification. We can only conclude that the lineup was “unnecessarily suggestive and conducive to irrep*188arable •• mistaken identification.’.’ (Stovall v. Denno (1967) supra, 388 U.S. 293, 302 [18 L.Ed.2d 1199, 1206]), and we hold that its grossly unfair makeup deprived defendant of due process of law. (Stovall v. Denno, supra.)
In reaching this determination, we do no more than recognize, as did the United States Supreme Court, that unfairly constituted lineups have in the past too often brought about th¿ conviction of the innocent. In United States v. Wade (1967) supra, 388 U.S. 218, 228-229 [18 L.Ed.2d 1149, 1158-1159], the Supreme Court observed: “A major factor contributing to the high incidence of miscarriage of justice from mistaken identification has been the degree of suggestion inherent in the manner in which the prosecution presents the suspect to witnesses for pretrial identification. A commentator has observed that ‘the influence of improper suggestion upon identifying witnesses probably accounts for more miscarriages of justice than any other single factor—perhaps it is responsible for more such errors than all other factors combined. ’ Wall, Eyewitness Identification in Criminal Cases 26. . . . And the dangers for the suspect are particularly grave when the witness’ opportunity for observation was insubstantial, and thus his susceptibility to suggestion the greatest. ’ ’ It was further observed (at pp. 229-230 of 388 U.S. [18 L.Ed.2d at p. 1159]) that once a witness picks the accused at the lineup he rarely retracts his identification and that in robbery eases the lineup presents “a particular hazard that a victim’s understandable outrage may excite vengeful or spiteful motives.”
The record in this case amply underscores the hazards of an improperly constituted lineup. By their own testimony the victims obtained no more than a fleeting glance at the robbery ear, and yet their in-court identification convinced the jury that defendant was guilty of the robbery. By their verdict the jurors disregarded defendant’s persuasive alibi, corroborated by three witnesses, that he spent the morning of February 23 at a business engagement more than 40 miles from the scene of the robbery. Indeed, had the jury accepted the alibi testimony in any respect it would have been required to reject Seeley’s in-court identification because, it will be recalled, he testified he had observed defendant in midmorning in the Robert Hall parking lot. On the other hand, the jury could have found that Officer Torres correctly recorded the license number and, therefore, that defendant or someone in his vehicle was in the vicinity of the crime at 1 p.m. The effect of Torres’ statement, however, was only to impugn *189defendant’s alibi. His testimony could not strengthen the prosecution’s case in chief because, as already explained, there was no evidentiary link between Torres’ observations and the crime. Other circumstances, moreover, lent credence to the defense: a police search of defendant’s home failed to reveal either the automatic pistol or the missing receipts. Despite intensive police interrogation defendant consistently maintained his innocence. No motive for the offense was ever established: defendant was steadily employed at a satisfactory wage and maintained a good reputation at work; he was married, has children, and was in no pressing financial difficulty. He had no prior criminal record in this state or in the state of his origin, Illinois. Finally, the accomplice was never charged; he apparently remains unknown to the police, and no known friends or associates of defendant have been subject to involvement.
The crucial circumstance remains that the victims identified defendant as the driver of the robbery car in a lineup so unfairly conducted that it deprived him of due process of law. Again, we find the opinion in United States v. Wade, supra, instructive: ‘ ‘ The trial which might determine the accused’s fate may well not be that in the courtroom but that at the pretrial confrontation, with the State aligned against the accused, the witness the sole jury, and the accused unprotected against overreaching, intentional or unintentional, and with little or no effective appeal from the judgment there rendered by the witness—-‘that’s the man.’ ’.’ (388 U.S. at pp. 235-236 [18 L.Ed.2d at p. 1162].) Here, too, defendant’s counsel could not free the defense from the taint of the improper lineup, and his client’s destiny was, in a real sense, sealed by the time the trial commenced.
In the event the People choose to try defendant again, we consider the rules to be applied to make certain that the retrial does not suffer from disability incurred through the illegal lineup.2 To overcome the effect of the taint, the People must now on voir dire show by clear and convincing proof that the in-eourt identifications were based upon observations of the accused at the scene of the robbery. *190(United States v. Wade (1967) supra, 388 U.S. 218, 240 [18 L.Ed.2d 1149, 1164].)3 “The phrase 1 *clear and convincing evidence’ has been defined as ‘clear, explicit, and unequivocal,’ ‘so clear as to leave no substantial doubt, ’ and ‘ sufficiently strong to demand the unhesitating assent of every reasonable mind.’ ” (In re Jost (1953) 117 Cal.App.2d 379, 383 [256 P.2d 71].) The prosecution will bear the burden of producing through the victims the requisite level of proof. The lineup served to enhance their memories so that they could identify defendant at trial. They now must totally eliminate from recollection all observations at the lineup and convince “every reasonable mind” that they distinctly recall defendant from their fleeting impressions during the robbery-. As suggested in Wade, this may be extremelly difficult.4 Upon retrial the victims can, of course, describe the robbery, but it is the identity of defendant as one of the robbers that presents grave problems on this record.
In the present case if the in-eourt identifications are not rehabilitated, the prosecution has no evidence to connect defendant with the offense-. While the People offered testimony that defendant’s ear was near the Robert Hall store at the time of the robbery, this without more would not establish *191proof beyond a reasonable doubt of his guilt. (See People v. Hall (1964) 62 Cal.2d 104, 112 [41 Cal.Rptr. 284, 396 P.2d 700].)
The judgment is reversed.
Traynor, C. J., Peters, J., Tobriner, J., Burke, J., and Sullivan, J., concurred.
On this point Robert Butkus testified as follows: "Q. And did you in that lineup see Mr. Caruso? A. Tes, I did. Q. Was there anyone else in that lineup that looked like him? A. That looked like him? Q. Tes. A. No. I would have to say no.” Olen Seeley gave this description: ”Q. . . . you say you went to a police station and saw Mr. Caruso in a lineup? A. Tes. Q. And how many other people were in that lineup? A, About four more. Q. Anyone else who looked like Mr. Caruso? A. No.” Officer Allen, who conducted the investigation, was unable to recall the makeup of the lineup. ”Q. And thereafter you were present, Sergeant, at some sort of a lineup "where Mr. Caruso was present? A. Yes, sir. Q. And do you recall how many other people were in that lineup? A. Besides Mr. Caruso? Q. Tes, sir. A. I believe there were four. Q. Isn’t it a fact, Sergeant, that there was no one else in that lineup who was actually of the same height, weight, and complexion as this defendant? A. No, that is not a fact. Q. That is not a fact? A. Not your whole statement. Q. There was nobody in that lineup who was as dark complected as Mr. Caruso? A. I would say, as I recall, probably not. Q. Do you remember, actually, who was in that lineup? A. No, I don’t. Q. Do you have any record as to who was there that looked like defendant? A. I don’t know that anybody else looked like this defendant.” Finally, defendant had this recollection: “Q. Was there anyone else in that lineup who was of your approximate size and weight, with dark features and dark hair? A. My size and weight with dark features and dark hair? Q. Tes. A. No. Definitely not.”
Since the in-eourt identifications here clearly-' and directly resulted from the illegal lineup, the People áre hot entitled" tb: a mere vacation of the proceedings -to determine' -in án evidentiary hearing whether the identifications:had .an 'independent-source. (See Gilbert v. California (1967) supra, 388 U.S. 263, 272-273 [18 L.Ed.2d 1178, 1186-1187]) From the factual account of this case it is obvious that the tainted imeourt *190identifications were essential to defendant’s conviction, and therefore the admission of this testimony constituted prejudicial error. (Chapman v. California (1967) 386 U.S. 18, 24 [17 L.Ed.2d 705, 710-711, 87 S.Ct. 824].)
We have previously said that " [t]he manner in which the lineup was conducted affects only the weight of the witnesses’ testimony, not its admissibility.” (People v. Parham (1963) 60 Cal.2d 378, 380 [33 Cal.Rptr. 497, 384 P.2d 1001].) But unless the "clear and convincing evidence ’ standard is satisfied in the context of a denial of due process, as in the instant case, the in-court identification testimony is now inadmissible. To the extent that Parham and similar eases are inconsistent with this rule (see, e.g., People v. Smith (1963) 223 Cal.App.2d 388, 393 [35 Cal.Rptr. 731], and People v. McLaine (1962) 204 Cal.App.2d 96,104 [22 Cal.Rptr. 72]), they can no longer be considered authoritative.
See United States v. Wade (1967) supra, 388 U.S. 218, 247-248 [18 L.Ed.2d 1149] (Black, J., dissenting in part and concurring in part) : "The ‘taint’—‘fruit’ determination required by the Court involves more than considerable difficulty. I think it is practically impossible. How is a witness capable of probing the recesses of his mind to draw a sharp line between a courtroom identification due exclusively to an earlier lineup and a courtroom identification due to memory not based on the lineup ? What kind of ‘ clear and convincing evidence ’ can the prosecution offer to prove upon what particular events memories resulting in an in-eourt identification rest? How long will trials be delayed while judges turn psychologists to probe the subconscious minds of witnesses ? All these questions are posed but not answered by the Court’s opinion.”
See also, -Ruffin, The Tainted Witness (1967) 15 U.C.L.A. L.Rev. 32, 40.