delivered the opinion of the Court.
In this tort action, Pamela Banks sued the City of Richmond and others1 for damages resulting from an explosion of natural gas in her apartment at the Jefferson Village Apartments in Richmond. Banks proceeded solely against the City in a jury trial. At the conclusion of plaintiffs case, the trial court granted the City’s motion to strike. The trial court was of the view that the evidence failed to establish that the City had done anything “wrong that proximately caused the accident.”
On appeal, Banks contends that the trial court erred in the following three particulars:
(1) in failing to hold that the City had a legal duty to make repairs of defective gas appliances and pipes within the consumer’s possession;
(2) in failing to hold that the City had a legal duty to cut off the supply of gas to the dwelling to which it supplied gas once the City was aware that there was a *132leak in the gas appliances or the pipes owned by the consumer; and
(3) in failing to resolve any doubts regarding sufficiency of the evidence in favor of the plaintiff and by assuming facts not in evidence.
The City’s main contention is that its acts or omissions did not proximately cause the explosion. We agree with the City; therefore, we assume without deciding that the City was negligent.
Banks moved into the apartment shortly after July 17, 1981. The explosion occurred on July 29, 1981. However, the bulk of Bank’s evidence was based on events that occurred prior to her occupancy.
Charlotte Dickerson lived in the apartment prior to Banks. Dickerson moved out in mid-June 1981. In September 1980, Dickerson began complaining of the smell of gas in her apartment. From late 1980 until February 1981 she made approximately nine telephone calls to the City, which supplied natural gas to her apartment complex, and to the rental office at Jefferson Village in which she complained of the problem. Dickerson said that four or five times she pointed out the furnace and the oven to the City employee who responded to her calls. Each time, the City employee would check those areas and then tell her to call her rental office.
On cross-examination, Dickerson clarified her testimony, explaining that of her several complaints concerning the smell of gas, only one complaint was made concerning her oven. All the other complaints centered on her gas furnace. She said that the only complaint concerning the oven occurred in mid-February 1981. The City employee who checked her oven at that time told her she had a small gas leak and told her to tell her apartment maintenance department.
Dickerson followed the City employee’s advice. The apartment maintenance man came to Dickerson’s unit later on the same day. When he checked the oven, he was able to smell gas but he could not find a leak. As a result, he turned off the gas to the oven.2
*133The gas to the oven remained turned off for as long as Dickerson remained in the apartment. Jefferson Village promised to replace the oven but this was never done. Once the gas to the oven was turned off, Dickerson did not smell any more gas. During her remaining time in the apartment, Dickerson had no further complaints concerning the smell of gas.
After Dickerson moved out but before Banks moved in, Bobby Rae Parker, a City gas and water service specialist, inspected the meter on the outside of the unit as part of his duties in transferring the service to Banks. It was Parker’s job to turn gas on and off at the meter.
In direct examination, Parker was asked whether on the day he went to the apartment he knew there was a “current problem” at the unit. He said that on that day he did not know there was a current problem. He was then impeached by deposition testimony where he had said, “I was told by someone else that there was a current problem.” After being impeached, Parker testified as follows: “When I left the apartment I knew there was a current problem there.” He went on to say that on the day of his visit, “[t]he cleaning person told me there had been a problem of some kind there around the oven, and I said, I will report it.” Parker said further that on the day of his visit he did not smell any gas while standing in the doorway to the apartment talking to the cleaning man. He also said that he did not inspect inside the apartment for a leak. He said he looked at the meter and observed that the gauges were operating normally. Parker explained that had the gauges been moving fast that would have suggested trouble. The gas was on when Parker arrived. He did not turn it off. Nor did he notify anyone other than the rental office of his conversation with the cleaning man.
At the rental office, Parker told a woman what the cleaning man had told him. The woman wrote down the message. However, Parker did not read what she wrote down. On a form that Parker filed with the City, he noted that office maintenance was checking on the oven.
Jeanette Agee worked at the rental office. On July 17, 1981, she wrote a note concerning the unit in question. The note read as follows: “Gas smell. Check oven.” She did not remember who gave her the information. She said she placed the note in the slot for the maintenance man, whose job it was to make repairs.
*134When Banks moved into the apartment, the oven did not work. She complained daily to the rental office. The rental office advised her that someone would fix the oven.
The explosion occurred on July 29, 1981. On that day, Harry Stahl, a maintenance man for Jefferson Village, was working at the apartment replacing oven and refrigerator racks. While Stahl was at the apartment, Banks told him that the oven did not work. She asked Stahl whether he could repair it. Stahl went to the furnace room and turned the gas on to the oven with a pair of pliers. Then he lit the oven’s pilot light. Stahl then left the unit to get materials he needed for other work in the unit. When he returned he smelled “[a] faint odor of gas.” Stahl went to the furnace room and turned off the gas to the oven. He did other work in the unit for “[a] couple of minutes” then returned to work on the oven. When Stahl went back to the oven he said he could no longer smell gas. Next, he “took the front cover off to the thermostat to the oven, and opened the two doors to the oven, and lit a cigarette lighter.” He testified that he lit the lighter to look inside the oven. The explosion followed instantly. He was blown through the air, across the kitchen, and through the kitchen cabinets.
Bernard D. Roberts, the Chief Utility Engineer for the City, testified concerning the cause of the explosion. He said that his investigation revealed “a rather large leak in the gas flexible connecting tubing” that connected to the back of the oven. In his opinion, the source of the fuel which caused the explosion was the oven area. He said when he turned on the gas to the oven he heard a “hissing sound” coming from the flexible connecting hose. According to Roberts, prior to the explosion the connecting hose was enclosed in the wall behind the oven.
Banks contends that the City is liable for the explosion because it did not turn off the gas at the meter on July 17, 1981, when the cleaning man told the City employee that there was a “current problem” concerning gas service to the apartment. Banks says the failure to turn off the gas, at the meter, was negligence on the part of the City. Banks argues further that had the gas been turned off at the meter, the explosion could not have occurred no matter how many times Stahl used a cigarette lighter to inspect the inside of the oven. Thus, according to Banks, the failure to turn off the gas at the meter was a proximate cause of the explosion. We disagree.
*135The definition of proximate cause is familiar. In Coleman v. Blankenship Oil Corp., 221 Va. 124, 131, 267 S.E.2d 143, 147 (1980), we said that:
“The proximate cause of an event is that act or omission which, in natural and continuous sequence, unbroken by an efficient intervening cause, produces the event, and without which that event would not have occurred.”
(Quoting Beale v. Jones, 210 Va. 519, 522, 171 S.E.2d 851, 853 (1970)). In Huffman v. Sorenson, 194 Va. 932, 937, 76 S.E.2d 183, 187 (1953), we defined proximate cause in slightly different terms; there we wrote as follows:
the proximate cause of an injury is that act or omission which immediately causes or fails to prevent the injury; an act or omission occurring or concurring with another, without which the injury would not have been inflicted. . . .
In Winfree v. Jones, 104 Va. 39, 43, 51 S.E. 153, 154 (1905), we explained that “the term ‘proximate’ excludes the notion of the intervention of any other culpable and efficient agency between the defendant’s dereliction and the loss.” In the early case of Connell v. Chesapeake, &c. R. Co., 93 Va. 44, 59, 24 S.E. 467, 469 (1896), we adopted the view that though a person is answerable for the consequences of a fault which are natural and probable, if the fault happens to concur with something extraordinary and not likely to be foreseen, the person will not be answerable.
The more difficult problem is to apply the rules relating to proximate cause to the facts of a particular case. We have stated that “[e]ach case necessarily must be decided upon its own facts and circumstances.” Huffman, 194 Va. at 937, 76 S.E.2d at 186. Similarly, in Scott v. Simms, 188 Va. 808, 816, 51 S.E.2d 250, 253 (1949), we said that “[t]here is no yardstick by which every case may be measured and fitted into its proper place. In each case the problem is to be solved upon mixed considerations of logic, common sense, justice, policy and precedent.” Further, in Spence v. American Oil Co., 171 Va. 62, 74, 197 S.E. 468, 473 (1938), we noted that in resolving the question of proximate cause one of the key questions is whether there exists an intermediate cause disconnected from the primary fault that, self-operating, caused the injury.
*136When we apply the foregoing considerations to the facts of this case it is clear that the proximate cause of this explosion was not the City’s failure to turn off the gas at the meter on July 17, 1981. If that was a cause at all, it was a remote cause.
The proximate cause of this explosion was the manner in which the maintenance man undertook to determine what was wrong with the oven. It was Stahl’s negligence that immediately caused the explosion. Absent Stahl’s misconduct, there would not have been an explosion.
Banks complains that had the gas been cut off at the meter there would have been no explosion. Yet, that simplistic “but for” argument does not resolve the proximate cause question. In Wyatt v. Telephone Company, 158 Va. 470, 163 S.E. 370 (1932), we rejected a similar argument that because a telephone pole was negligently placed it was the proximate cause of a collision with the pole. We said that the latter did not follow from the former. Banks’ approach overlooks the fact that when Stahl was asked whether he could fix the oven the gas supply to the oven was off and had been off for months. The gas was purposefully turned on as part of Stahl’s effort to find out what was wrong.
In checking on the oven, Stahl took no steps to examine the oven before he turned on the gas. He did not contact the maintenance office or the rental office to determine whether there was a reason for the oven’s being turned off. He simply turned on the gas to the oven, smelled the faint odor of gas in the kitchen, then turned the gas to the oven off. Thereafter, he placed an open flame in a confined area where minutes before he had certain knowledge of the presence of gas.
It is not reasonably foreseeable that a maintenance man will search for a gas leak with fire. Given the procedure used by Stahl, the fact that the gas was not turned off at the meter is not legally significant. Stahl obviously was of the view that to determine what was wrong with the oven he had to see what happened when he turned on the gas. Stahl’s conduct so entirely superseded that of the City that it alone produced the injury. Stahl’s conduct amounted to an intervening independent act that affected and was the immediate cause of the injury. See Huffman, 194 Va. at 939, 76 S.E.2d at 188. Given the actual facts surrounding the explosion, the City’s failure to turn off the gas at the meter was nothing more than a mere circumstance of the explosion. See Hubbard v. Murray, 173 Va. 448, 3 S.E.2d 397 (1939).
*137In Spence, we stated that the extraordinary manner in which harm occurs may prevent the primary actor’s conduct from being the proximate cause of an event. We relied upon Comment (a) to the Restatement of Torts § 435 (1934). That comment provides in pertinent part as follows: “[T]he manner in which the harm occurs may be so highly extraordinary as to prevent the actor’s conduct from being a substantial factor in bringing it about.” Spence, 171 Va. at 76, 197 S.E. at 474. This is such a case. What Stahl did was “highly extraordinary.” It completely overshadows anything that the City did.
Normally, questions of proximate cause are for the jury. However, where the facts are not in dispute and reasonable men could not differ as to the inferences to be drawn from the facts, the issue can be resolved as a matter of law. See Hubbard v. Murray, 173 Va. at 457, 3 S.E.2d at 402. There was no dispute here concerning Stahl’s activities. We do not think that reasonable men could differ concerning the facts and the inferences to be drawn therefrom. Consequently, we will affirm the trial court’s judgment that struck plaintiff’s evidence.
Affirmed.