The defendant, Daniel F. Rollins, appeals from his conviction by a District Court jury of operating a motor vehicle while under the influence of intoxicating liquor in violation of *695G. L. c. 90, § 24(1)(a)(1).1,2 He asserts that there was error in two respects. First, he claims that his pretrial motion to suppress the result of a breathalyzer test should have been allowed because the breathalyzer machine used had automatically become decertified by virtue of an earlier invalid test. Second, the defendant contends that the prosecutor’s closing argument was improper. He takes issue with the prosecutor’s assertion in closing that a breathalyzer test result of .09 percent is inconsistent with the defendant’s claim that he had only two or three beers that day, where there was no evidence demonstrating what such a test result indicated with respect to consumption of alcohol by a given individual. Because the evidence shows that the earlier breathalyzer test produced an invalid result not because of a mechanical defect but because of operator error, we conclude that the machine was not legally decertified. We conclude further that the prosecutor’s statement was supported by a permissible inference from the evidence, and was therefore within the bounds of proper argument. Accordingly, we affirm the judgment.
1. Admission of breathalyzer test result. We review first the judge’s denial of the defendant’s motion to suppress the result of a breathalyzer test administered to him at a police station. The judge’s findings of fact in connection with the motion were supported by the evidence at the motion hearing, and we therefore defer to those findings. See Commonwealth v. James, *696427 Mass. 312, 314 (1998). The judge found that, on the evening of March 17, 2002, Officer Sean Frost of the Topsfield police department observed the defendant stop at a traffic light at the intersection of Boston Street and Ipswich Road. When the light turned green, the defendant crossed the center Une and, at a high rate of speed, passed other cars stopped at the intersection. Officer Frost pulled the defendant over and, on the basis of his observations and the defendant’s failure of two of three field sobriety tests, concluded that the defendant was operating the motor vehicle while under the influence of alcohol. Accordingly, Officer Frost arrested the defendant and transported him to the Topsfield police station.
At the police station, Officer Frost advised the defendant of his rights, including his right to a breathalyzer test, which the defendant agreed to take. Officer Frost, a certified breathalyzer operator, administered the test himself. The breathalyzer machine that he operated was the police department’s Intoxilizer 5000, an approved breath-testing device that had passed its annual certification inspection and contained unexpired simulator solution. See 501 Code Mass. Regs. §§ 2.38, 2.39, 2.41 (1996).
Before a breathalyzer test can be administered on the Intoxilizer 5000, the machine must be warmed up so that the simulator solution, which is used to calibrate the machine, reaches the appropriate temperature. A light on the Intoxilizer indicates when the equipment is sufficiently warmed up and is ready to be used. Once this occurs, a suspect blows into the machine while a tone sounds. The machine thereupon registers both the suspect’s blood alcohol level and the level of the simulator solution. The measurement of the simulator solution is referred to as a calibration standard analysis; that measurement must be between .14 and .16 percent in order for the machine to be calibrated correctly. If so, the suspect blows into the machine again, and the machine issues a second blood alcohol level reading. Assuming that the two blood alcohol level readings do not differ by more than .02 percent,3 the lower of the two readings is the determination of the suspect’s blood alcohol level. See 501 Code Mass. Regs. § 2.56 (1996).
*697Officer Frost administered the breathalyzer test to the defendant after the machine had been permitted to warm up for fifteen to twenty minutes, but before the machine’s ready light had come on. When the defendant first blew into the machine, his blood alcohol level registered .10 percent, but the calibration standard analysis was .13 percent, indicating that the equipment was not calibrated correctly. After consulting with Officer Hayward, who was also present at the police station, Officer Frost let the machine warm up further; the ready light came on, and he administered the breathalyzer test a second time. The defendant’s blood alcohol level again registered .10 percent, but on this occasion the simulator solution registered a calibration standard analysis reading of .14 percent, indicating that the equipment was now calibrated accurately. Officer Frost tested the defendant’s blood alcohol level again, this time obtaining a reading of .09 percent. It was this final reading that was ultimately admitted in evidence.
The defendant challenges the admission in evidence of this final result on the ground that the breathalyzer machine had, by reason of applicable regulations, been rendered decertified when a test producing an incorrect calibration was conducted. He argues that suppression is required because the simulator solution reading of .13 percent on the initial attempt had the effect of decertifying the breathalyzer, rendering any subsequent test results produced by that equipment invalid, and thus inadmissible in evidence, until the machine was recertified. See G. L. c. 90, §§ 24, 24K; 501 Code Mass. Regs. § 2.03 (1996) (no breath test valid for purposes of enforcement of G. L. c. 90 unless administered by certified breath-test operator, using certified equipment, and operating in accordance with procedures established under 501 Code Mass. Regs. §§ 2.00 et seq.).
More particularly, the defendant relies on those provisions of the regulations that govern the decertification of equipment. Breathalyzer equipment must be certified annually by the office of alcohol testing4 on the basis, among other things, that the equipment has successfully completed a certification testing program. See 501 Code Mass. Regs. § 2.39 (1996). The equip*698ment is, however, subject to an additional testing requirement, specifically, that “every calibration standard analysis of a breath testing device, as conducted pursuant to 501 [Code Mass. Regs. §] 2.56, shall be deemed to be a test of such device.” 501 Code Mass. Regs. § 2.41(1) (1996). This is followed in the applicable regulations by a provision that “[a] certified breath testing device or simulator which fails to successfully complete the certification testing program established under 501 [Code Mass. Regs. §] 2.40 or the periodic testing program established under 501 [Code Mass. Regs. §] 2.41 shall be decertified by the Office of Alcohol Testing.” 501 Code Mass. Regs. § 2.42 (1996). Because, the defendant’s argument continues, periodic testing under § 2.41(1) includes by definition the calibration standard analysis that takes place in every test under § 2.56, a single failure of the equipment in that respect automatically brings about decertification under § 2.42 and thereby renders any result obtained after decertification inadmissible in evidence. Here, he contends, the first calibration standard analysis reading of .13 percent constituted a failure to satisfy periodic testing requirements, thus invalidating any further readings taken at that time.
We agree with the defendant’s interpretation of the governing regulations, but not with his application of those regulations to the facts of the case. Periodic testing of breathalyzer equipment is contemplated by G. L. c. 90, § 24K, and is required by 501 Code Mass. Regs. §§ 2.39 and 2.41, so as to ensure accuracy. See Morris v. Commonwealth, 412 Mass. 861, 866-867 (1992). “The . . . regulations require the machines to be tested and proven every time they are used.” Id. at 866. In other words, should the equipment fail on any given occasion, it cannot be used to generate admissible test results until the defect is rectified and the machine is recertified.
It is, however, the failure of the equipment, not the failure of the operator, that is the target of this portion of the regulations. The inclusion of every calibration standard analysis as a test that must be passed in order to avoid decertification presupposes that such analysis will take place in the context of a correct use of the machine. Here, the motion judge found in essence that the machine had not been given sufficient time in *699which to warm up; that there was nothing about the equipment itself that malfunctioned; and that, once given sufficient warm-up time, the breathalyzer functioned properly in all respects. These findings were supported by the evidence and provided a permissible basis for allowing the breathalyzer test result in evidence. The defendant was free to challenge the reliability of the test result at trial through cross-examination or otherwise, including by reference to the initial calibration standard analysis reading. He was not entitled either on these facts, or on a reasonable interpretation of the regulations in light of their purpose, to have the result excluded.
2. Closing argument. At the close of the Commonwealth’s final argument, the defendant objected to remarks of the prosecutor to the effect that the breathalyzer blood alcohol level reading of .09 percent was inconsistent with the defendant’s testimony that his alcohol intake consisted of only three beers consumed two to three hours before he was stopped. The challenged statement was the following:
“[Ajfter you hear what the law is, you will recognize what the significance of a .09 is, and that’s important because it is entirely, entirely inconsistent with what the defendant says he consumed for alcohol that evening. If you’re to credit his testimony that he consumed — Well, let’s say this. First he tells the officer he only consumed two beers. Then he says three beers. So if you’re to believe that he consumed three beers, and he tells the officer he consumed three beers two to three hours before he was stopped, and consumed two to three beers two to three hours before he was stopped and had food — I believe it was a chicken dinner — it’s inconsistent with the fact that the officer still smelled a moderate odor of alcohol on his breath, his eyes were still bloodshot and glassy, and he blew a .09 on a breathalyzer when he was stopped. That is inconsistent with one who has consumed only two to three beers.”
It is the defendant’s contention that no evidence justified the prosecutor’s argument that three beers consumed two to three hours earlier could not have produced a breathalyzer blood alcohol level reading of .09 percent, and that he was therefore prejudiced by an unsupported inference that his evidence that *700he was not operating the motor vehicle while under the influence of alcoholic beverages could not be true.
The evidence on the issue whether the defendant was impaired by alcohol while operating a motor vehicle on the occasion in question was in conflict. The Commonwealth relied on the breathalyzer test result (including the permissible inference therefrom that the defendant was in fact under the influence), together with Officer Frost’s testimony regarding his observations of the defendant and the defendant’s performance of field sobriety tests. The defendant countered with his own testimony that he had consumed three beers that day at a party for his grandmother; that his eyes may have appeared bloodshot because of an eye condition; and that he failed field sobriety tests because of an ankle injury. The defendant’s brother and uncle also testified that he did not appear intoxicated at the party or while driving home after and that the defendant’s brother had spilled beer on the truck’s floor earlier in the afternoon, thus causing the truck to smell of alcohol. Neither the Commonwealth nor the defendant offered expert or other evidence as to the following: the significance of a .09 percent breathalyzer reading; how many drinks would be consistent with a .09 percent breathalyzer reading for a person of the defendant’s age and size; or a likely blood alcohol level generated by the consumption of three beers in two to three hours by a person of the defendant’s age and size.
As indicated, the case was tried under a prior version of the statute that defined the offense as operation of a motor vehicle “while under the influence of intoxicating liquor,” G. L. c. 90, § 24(l)(o)(l), as amended through St. 2002, c. 52, § 2, and that permitted the jury to infer that there was operation while under the influence if the percentage of alcohol in the defendant’s blood was .08 percent or higher, see G. L. c. 90, § 24(l)(e), as amended through St. 1996, c. 151, § 236. That inference, however, was permissible, not mandatory, and a defendant could overcome the effect of the breathalyzer test if he could convince the jury by other evidence that he was in fact not alcohol-impaired. See Commonwealth v. Mahoney, 400 Mass. 524, 532 & n.2 (1987). The defendant in the present case attempted to do so.
*701Because the inference from the breathalyzer test was permissive only, the jury were free to ignore the test result even in the absence of other evidence, see Commonwealth v. Moreira, 385 Mass. 792, 795 (1982). Where the defendant offered evidence that he was in fact not operating a motor vehicle while under the influence, the prosecutor was certainly entitled to argue the Commonwealth’s version of the case. Such argument obviously must be limited to facts supported by evidence or reasonable inferences therefrom. See Commonwealth v. Coren, 437 Mass. 723, 730 (2002), citing Commonwealth v. Good, 409 Mass. 612, 623 (1991). Within those parameters, the prosecutor may argue forcefully why the jury should credit the Commonwealth’s evidence and draw the inferences that the Commonwealth seeks, while rejecting the evidence and inferences proffered by the defendant. See Commonwealth v. Kozec, 399 Mass. 514, 516 (1987); Commonwealth v. Shea, 401 Mass. 731, 738-739 (1988).
Here, the prosecutor was armed with a breathalyzer test result of .09 percent and the accompanying permissible inference that the defendant was under the influence of intoxicating liquor. It was well within the scope of proper argument to ask the jury to credit the scientific test buttressed by police observation rather than the conflicting testimony of the defendant and his relatives. We are not entirely convinced that the statement by the prosecutor (that the defendant’s breathalyzer result of .09 percent was inconsistent with his claim that he had consumed only three beers two to three hours before he was stopped) did not cross the line into territory that required expert testimony. While a similar argument was upheld as based on a reasonable inference in Commonwealth v. Howe, 405 Mass. 332, 335-336 (1989), we observe that that case involved a breathalyzer result of .18 percent compared to the then permissible inference of operation of a motor vehicle while under the influence from a breathalyzer reading of .10 percent or more. See G. L. c. 90, § 24(1)(e), as appearing in St. 1980, c. 383, § 1. The present case features a breathalyzer result of .09 percent compared to a permissible inference based on a reading of .08 percent, thus rendering the inference that the defendant misrepresented his drinking considerably less compelling as a reasonable application of the jury’s common sense and general understanding. Be that as it *702may, we are satisfied that, if there were error, it was not prejudicial on this record. There is little likelihood that the jury’s ability to understand and evaluate the evidentiary conflicts was affected.
Judgment affirmed.
The defendant was also charged with two civil motor vehicle violations. The judge found him not responsible, apparently on technical grounds.
The alleged offense occurred on March 17, 2002. At that time, G. L. c. 90, § 24(1)(a)(1), as amended through St. 2002, c. 52, § 2, criminalized, in pertinent part, the operation of a motor vehicle on a public way “while under the influence of intoxicating liquor.” General Laws c. 90, § 24(1)(e), as amended through St. 1996, c. 151, § 236, in turn provided that, should there be evidence that the percentage, by weight, of alcohol in the defendant’s blood was “eight one-hundredths or more, there shall be a permissible inference that such defendant was under the influence of intoxicating liquor.” During the following year, the statute was amended to make criminal the operation of a motor vehicle on a public way “with a percentage, by weight, of alcohol in [the operator’s] blood of eight one-hundredths or greater, or while under the influence of intoxicating liquor.” St. 2003, c. 28, § 1, effective June 30, 2003. The statute’s earlier reference to a permissible inference based on such evidence was deleted. See St. 2003, c. 28, § 4, effective the same date. Given the date of the alleged offense, the earlier version of the statute is applicable.
If they differ by more than .02 percent, the test is considered not valid.
The office of alcohol testing is a unit within the Department of State Police. See 501 Code Mass. Regs. § 2.36 (1996).