The appellant was convicted of four counts of aggravated robbery and sentenced to thirty years on each count, the sentences to run consecutively. The only argument made on appeal is that the trial court was wrong in permitting the victims to make in-court identifications, because they were inherently unreliable. We find no merit in this argument, and Banks’ conviction and sentence are affirmed.
The four counts of aggravated robbery were made because Banks and another man robbed four women about 11:30p.m. on July 14, 1983, near the Excelsior Hotel, Little Rock, Arkansas. The women left the hotel and, as they were approaching their vehicle that was parked on a nearby street, they saw two men sitting on the curb in front of the car. As they reached the car, the men pulled bandanas over the lower part of their faces and one of the men grabbed one of the victims, pushed her into the car on top of another woman, who was in the driver’s seat, and demanded their purses at gun point. The women finally surrendered their purses and the men fled on foot. The matter was reported to the police and it was investigated. A man named Johnny Henderson reported to the police that he was an eyewitness to the incident and was acquainted with the defendant Banks and identified him as being one of the men. The next day the women went to the police station to make a statement and view photographs of possible suspects. The defendant’s picture was not among the photographic spread. One victim *286thought she could identify one of the men. The women could not give the police very accurate descriptions of the men who robbed them. Six days after the robbery the women were asked to view a lineup. Each woman viewed the lineup separately and three of them made positive identifications of Banks as being one of the men who robbed them. There was testimony that one of the police officers told one of the victims that she had identified a man who was also picked by an eyewitness. There is no evidence that the police prompted the victims in any way before they viewed the lineup.
Whether an in-court identification meets constitutional standards is essentially for the trial court to decide. We reverse the trial judge only if his decision is clearly erroneous. Kellensworth v. State, 278 Ark. 261, 644 S.W.2d 933 (1983); Perry v. State, 277 Ark. 357, 642 S.W.2d 865 (1982). The question is whether or not the in-court identification has been tainted impermissibly by a previous suggestive lineup. We cannot say in this case that the court was clearly wrong in allowing the in-court identification to go to the jury. There is nothing in our judgment to suggest that the lineup was so unfair as to taint the later in-court identification nor that the trial court was wrong in ruling that the in-court identification was reliable. The complaint as to the lineup is largely to the officer’s remark to one of the women that she had identified the same man an eyewitness did. The testimony was conflicting as to whether or not this statement was made. We have held that such a comment, if it was in fact made, is improper but, when made after the selection, is not prejudicial. Perry v. State, supra. Furthermore, even when a pretrial procedure is improper, the in-court identification is improper only if, under the totality of the circumstances, the judge determines there is a likelihood of misidentificadon. Perry v. State, supra. Factors for determining reliability of the in-court identification include whether the victims had sufficient opportunity to view the defendant, the time lapse between the crime and the identification, the level of certainty of the identification, and the accuracy of the prior description of the criminal. Hogan v. State, 280 Ark. 287, 657 S.W.2d 534 (1983).
The main argument of the appellant as to reliability is *287that the women gave poor descriptions arid they were not in a position to view these defendants when they were robbed because most of them paid no attention to the men sitting on the curb until after they had raised the bandanas over a portion of their faces. Of course, people do not expect to be robbed and the descriptions by victims are often vague and indefinite. However, these victims were in a position to view their assailants; three of them said they could positively identify Banks and did so in a lineup and later in court. There is no reason to disturb the trial court’s ruling and the judgment is affirmed.
Affirmed.