OPINION
MANDERINO, Justice.Appellant Hugh Fant, a black male, was tried before a judge and jury and convicted of two counts of murder of the third degree in connection with the shooting deaths of two men in a Uniontown nightclub. Post-verdict motions were denied, and appellant was sentenced to two terms of imprisonment of 10 to 20 years, the terms to run consecutively. This direct appeal followed.
Appellant raises five assignments of error, only one of which we need address here. Appellant claims he was denied due process of law when the trial court refused his request to sequester the prosecution’s witnesses, some of whom placed Fant at the scene of the killings and others who testified they witnessed Fant actually do the shooting. We agree with appellant that under the circumstances of this case, in which appellant was never identified at a *589pretrial lineup, photographic lineup, showup, or other pretrial identification procedure, the trial court’s refusal to sequester witnesses was an abuse of discretion which dictates that appellant be granted a new trial.
According to prosecution witnesses, several men were involved in a minor skirmish inside the Wishing Well Lounge in Uniontown, Pennsylvania at approximately 1:30 a.m. on December 15, 1974. Three of the men, whom prosecution witnesses identified at trial as appellant and two companions, were ejected from the nightclub. Another altercation then took place in the parking lot. The nightclub’s owner, Bruce Fisher, sprayed some mace into the scuffling group in order to restore order. The three men left by automobile, vowing to return.
Shortly thereafter four men returned. Two of the men were black and two were white. One was brandishing a rifle; one carried a handgun. The four men entered the nightclub through the front door. Fisher approached the four and was struck on the head with a beer bottle. As Fisher turned and began to walk away, one of the men shot him in the back. The same man then fired several more shots, killing one of the nightclub’s patrons.
Appellant was arrested the following morning. Between the time of appellant’s arrest and trial — approximately nine months — no identification procedures were utilized in an effort to have any of the prosecution’s witnesses identify appellant as the person responsible for the killings at the Wishing Well Lounge. The prosecution has not offered an explanation for its failure to utilize a photographic array, to conduct a lineup or showup, or to make any other efforts to obtain an identification of appellant prior to the initiation of judicial proceedings. At the preliminary hearing, however, the record indicates that the only witnesses who testified stated that they could not identify appellant as the killer.
Rather than make an effort to have the prosecution witnesses identify appellant via some type of objective pretrial identification procedure, the prosecution chose to have these witnesses identify appellant for the first time in open *590court. Because defense counsel’s motion to have the witnesses sequestered was denied, appellant was repeatedly identified by witnesses who had heard numerous other witnesses testify during the trial that appellant was present at the Wishing Well Lounge or was the one who fired the fatal shots. These witnesses identified appellant for the first time not from a lineup including other black males of similar physical characteristics, but while he was seated at counsel's table in open court, the only black man present, and after he had been charged with two murders. The suggestiveness inherent in these identifications of appellant is, in our view, palpably clear.
Our analysis of the present appeal must start with basic precepts fundamental to our system of criminal law. A criminal trial, at the very least, should aim at truthful fact-finding, including accurate eyewitness identifications. This is true whether the identification occurs at a pretrial lineup or within the four walls of the courtroom. For obvious reasons, all of society has an interest in accurate criminal identifications. To this end, we have required that police administrators and prosecutors adopt procedures that will give fair assurance against the dreaded consequences of misidentification. A review of this case finds such assurances completely wanting.
First, each witness identifying appellant at trial did so by way of a one-on-one confrontation. The practice of showing a suspect singly for the purposes of identification has been uniformly condemned. See, e. g., Manson v. Braithwaite, 432 U.S. 98, 97 S.Ct. 2243, 53 L.Ed.2d 140 (1977); Stovall v. Denno, 388 U.S. 293, 87 S.Ct. 1967, 18 L.Ed.2d 1199 (1967); Commonwealth v. Fowler, 466 Pa. 198, 352 A.2d 17 (1976); cf. ALI Model Code of Pre-Arraignment Procedure, §§ 160.1 and 160.2 (1975) (identification by use of a showup or the display of only a single photograph). The reason for this condemnation of one-on-one identifications is easily stated: An identification of a suspect from among other individuals of similar physical make-up is far more reliable, and far less suggestible than an identification when only the *591accused is shown to the witness. And, “reliability is the linchpin” in determining the admissibility of any confrontation identification. Manson v. Braithwaite, supra, 432 U.S. at 114, 97 S.Ct. at 2253, 53 L.Ed.2d at 154.
The problem of suggestiveness inherent in a one-on-one confrontation becomes more acute when the identification is made in open court. This fact was recognized in this Court’s plurality opinion in Commonwealth v. Fowler, supra, 466 Pa. 198, 203-204, 352 A.2d 17, 19-20 (1976):
“Trial testimony identifying one as the person observed at the time of a crime is a one-on-one confrontation involving circumstances even more suggestive than those present at pretrial one-on-one confrontations. During the trial, the identifying witness knows that the defendant present in the courtroom has been accused, arrested, and is being tried for the crime. Prior to trial, such circumstances may not yet have occurred or may not yet be known to the witness.”
In the present case, it was, of course, clear to everyone in the courtroom that appellant was the accused. The risk of misidentification is obviously much greater when an accused is picked out under these circumstances rather than from a photographic array or lineup.
It is against this background that we must examine the trial court’s denial of defense counsel’s request to sequester witnesses. It is true that a trial court’s decision regarding sequestration of witnesses is reversible only for a clear abuse of discretion. E. g., Commonwealth v. Hamilton, 459 Pa. 304, 311-12, 329 A.2d 212, 215-16 (1974), cert. denied, 420 U.S. 981, 95 S.Ct. 1411, 43 L.Ed.2d 663 (1975); Commonwealth v. Yount, 455 Pa. 303, 314 A.2d 242, 250 (1974); Commonwealth v. Kravitz, 400 Pa. 198, 218, 161 A.2d 861, 870 (1960), cert. denied, 365 U.S. 846, 81 S.Ct. 807, 5 L.Ed.2d 811 (1961). We think it was an abuse of discretion not to sequester witnesses under the facts of this case.
The court’s refusal to sequester witnesses only exacerbated the suggestiveness attending the in-court identifications *592of appellant. For example, the first prosecution witness to identify appellant as the killer — one Margaret Ruehmling— did so after twelve other witnesses had testified either that they saw appellant at the Wishing Well Lounge that night, or that they saw appellant approach the establishment with a rifle. The psychological pressures that influence a witness to positively identify someone after that witness has watched twelve others identify that same person as being at the scene can be overwhelming. The trial court’s decision would have been on a much different footing had some or all of the witnesses previously identified appellant at some type of fairly conducted, pre-trial identification procedure. In such a case, we have some assurance that the in-court identification is an accurate one, since it is an identification made apart from the compelling suggestibility of the trial setting.
Neither the trial court’s post-verdict opinion, nor the prosecution’s brief, has proffered any reason in support of the refusal to sequester witnesses. The trial court and the prosecution state only that many witnesses viewed appellant at the scene on the night of the killing. Therefore, it is argued, it was not an abuse of discretion to allow witnesses to hear other witnesses place appellant at the scene of the killings.
This justification for refusing to sequester the witnesses completely misperceives the purpose behind a trial court’s authority to sequester: to prevent one witness from shaping his testimony — in this case, the identification of appellant — in accordance with testimony given by other witnesses. See generally Commonwealth v. Turner, 371 Pa. 417, 88 A.2d 915 (1952); VI Wigmore on Evidence §§ 1837-1838 (Chadbourn rev. ed. 1976). Indeed, it is precisely because more than one witness will be testifying to a particular “fact” that a trial court is asked to sequester witnesses. It makes no sense to argue that it was a proper exercise of discretion to deny such a request because more than one witness testified to that “fact”. Where, as here, witnesses identifying appellant did so for the first time in the highly *593suggestive setting of an open court; where that suggestive identification was the first time in nine months that many of the witnesses were asked if appellant was the person they saw on the night in question; and where witnesses, after the first prosecution witness testified, heard anywhere from one to sixteen other persons identify the appellant, we must agree with appellant that his conviction, based almost entirely on eyewitness testimony, was so unfairly obtained as to violate due process of law. See, e. g., Stovall v. Denno, 388 U.S. 293, 299, 87 S.Ct. 1967, 1971, 18 L.Ed.2d 1199, 1205 (1967) (“it remains open to all persons to allege and prove . that the confrontation resulted in such unfairness that it infringed his [or her] right to due process of law”).
We are not unmindful that this senseless killing took place in a crowded nightclub in front of a number of people, many of whom testified for the prosecution in this case. Facially, therefore, it might be said that the in-court identifications of appellant had an independent basis, thus obviating the need to grant appellant a new trial. Compare Neil v. Biggers, 409 U.S. 188, 93 S.Ct. 375, 34 L.Ed.2d 401 (1972); Commonwealth v. Connolly, 478 Pa. 117, 385 A.2d 1342 (1978); Commonwealth v. Watkins, 239 Pa.Super. 144, 361 A.2d 656 (1976). We reject such reasoning in this case, for we deal here with more than simply whether or not the witnesses could make an accurate identification. We deal here with the fundamental fairness guaranteed by the fourteenth amendment’s due process clause. United States v. Lovasco, 431 U.S. 783, 97 S.Ct. 2044, 52 L.Ed.2d 752 (1977). There can be no doubt that identification procedures can be violative of due process, e. g., Stovall v. Denno, supra, Foster v. California, 394 U.S. 440, 89 S.Ct. 1127, 22 L.Ed.2d 402 (1969), and in our view, the procedure utilized in this case placed an intolerable taint on the very cornerstone of our criminal justice system, a fair trial. In such a case, a new trial must be granted, for every criminal defendant is entitled to at least that much. See Commonwealth v. Martinolich, 456 Pa. 136, 162-63, 318 A.2d 680, 695 (1974), and cases cited therein.
*594Lastly, we note that the prosecution’s brief seems to argue that appellant should not be permitted to challenge the trial court’s refusal to sequester because he did not move to suppress the identification evidence at a pre-trial suppression hearing. The prosecution cites no authority for this proposed requirement. The sequestration issue is separate from the identification issue. Because one does not move to suppress does not mean the right to sequestration is lost when sequestration is necessary to guarantee a fair trial. The issue of sequestration was raised at trial and in post-verdict motions. Under these circumstances appellant is not precluded from raising the issue in this appeal.
Judgment of sentence is reversed and a new trial granted.
EAGEN, C. J., and NIX, J., concurred in the result. POMEROY, J., filed a dissenting opinion in which O’BRIEN and LARSEN, JJ., joined.