Judgment rendered on July 10, 1961, convicting the defendant of the crime of murder in the second degree, unanimously reversed on the law and in the exercise of discretion, and a new trial ordered. At the trial the sole issue was whether or not the defendant was legally insane at the time of the commission of the crime. The only expert testimony on this issue was that of the psychiatrist called by the prosecution. The opinion of this witness as to the sanity of the defendant was based in part upon conversations had with other psychiatrists who had not testified at the trial and upon medical records, none of which had been admitted into evidence. Such material may not be considered by an expert witness in forming his opinion (People v. Samuels, 302 N. Y. 163, 172). The witness also based his opinion in part on his knowledge of the trial record. However, a witness may only base his opinion upon the testimony in the trial record if such testimony is included within a properly presented hypothetical question. (People v. Keough, 276 N. Y. 141, 145.) In this case, not only did the witness base his opinion upon the trial record — apart from the hypothetical question — but the hypothetical question itself was improper in form. A hypothetical question should be so framed as to enable the expert to tender his opinion based upon the assumption that the facts hypothetically stated to him are true (People v. McElvaine, 121 N. Y. 250; Matter of Bergman v. Mergenthaler Linotype Co., 274 App. Div. 553; Christastie v. Elmira Water, Light & R. R. Co., 202 App. Div. 270). In this case the question contained some assumptions of fact which were in conflict with other assumed facts. While it may well be that many of the defects regarding the hypothetical question were cured during the cross and redirect examination of the psychiatrist we are of the opinion that in the circumstances of this case and in the interest of justice, a new trial is required. Concur — Breitel, J. P., Rabin, Stevens, Steuer and Bastow, JJ.