This opinion is written pursuant to an appeal by the Commonwealth to the Superior Court as a result of the finding at a suppression hearing on August 21, 1990. At the hearing in question, this court found that certain breathalyzer test results were inadmissible based on the Commonwealth’s failure to recalibrate the breathalyzer or remove it from service as required by 67 Pa. Code §77.24.
The Pennsylvania Code, 67 Pa. Code §77.24, provides, in pertinent part, as follows:
“(b) Procedures. Alcohol breath tests shall be conducted by a certified breath test operator. Accuracy inspection tests and calibrations conducted using breath test equipment shall be performed by a certified breath test operator, the manufacturer or its authorized representative or a person who has received comparable training or instruction. Alcohol breath tests, accuracy inspection tests and calibrations conducted using breath test equipment shall be performed in accordance with accepted standard procedures for operation specified by the manufacturer of the equipment or comparable procedures. The procedures for alcohol breath testing shall include, at a minimum:
“(2) One simulator test using a simulator solution designed to give a reading of 0.10 percent, to be conducted immediately after the second actual alcohol breath test has been completed. The lower of the two actual breath test results will be the result used for prosecution. The test results will be disregarded, and the breath test device will be removed from service under section 77.25(b)(4) (relating to accu*591racy inspection tests for Type A equipment) if one of the following occurs:
“(i) If the difference between the results of the two actual alcohol breath tests is 0.02 or more, for machines read to the second decimal place, or 0.020 or more for machines read to the third decimal place.”
Defendant, George Mabrey, was pulled over at approximately 12:19 a.m. on February 23, 1990 in Bensalem, Bucks County, Pennsylvania. Officer Mark Blaszczyk, the patrolman on the scene, testified that defendant’s Lincoln Town car had swerved to the left several times and that defendant had a strong odor of alcohol on his breath, glassy eyes and slurred speech. The officer also testified that defendant failed some field sobriety tests and that he appeared to be under the influence of alcohol.
Defendant was transported to police headquarters and after consulting with his attorney, agreed to take a breathalyzer test. Officer Blaszczyk administered the test through use of the Intoximeter 3000. The test results were 0.16 percent blood alcohol content.
At the suppression hearing, the evidence showed that the breath-testing machine was given its last monthly accuracy test on February 9, 1990, approximately two weeks before defendant’s test. The evidence further indicated that the machine was calibrated in October 1989. The machine in question makes readings to the third decimal point. The evidence showed that there were at least seven known deviations greater than 0.020 on the machine from October 12, 1989 until the defendant was tested on February 23, 1990. These deviations occurred as follows.
On November 10, 1989, Officer Thomas Campel-lone had given a breathalyzer examination on the machine in question to a Mr. Bruce Vineland. The *592mouth alcohol readings were 0.221 and 0.245 percent, thus greater than the permissible 0.020 variation. Officer Campellone testified that he made a notation in the log book that the reason for this was that Vineland had given improper breath samples. The machine was not taken out of service or reca-librated. Instead, the subject was retested on the same machine and valid samples were received.
On December 7, 1989, Sergeant Jones administered a test on the same machine to a Mr. Decola. The first test showed a result of 0.212 percent but the second test was noted as “clear” in the log book. Officer Donald Schwab, the officer in charge of recalibration of the breathalyzer machines, stated that “clear” indicated that no second test result was obtained because the test aborted three times and the machine therefore shut down automatically. However, the same subject was retested after accuracy tests were performed on that machine and valid samples were received.
On December 18, 1989, Officer Schwab used the machine to test a Mr. Truex. In that test, there was a problem with the solution. While the first test registered correctly, the second test indicated a 0.090 percent. Officer Schwab testified that the jar in which the liquid was located was not screwed on tight. He took another set of tests which resulted in readings of 0.208 and 0.232 percent, once again yielding greater than the 0.020 variance. However, this was explained away by the officer who stated that Truex had given a poor sample by leaking air from the sides of his mouth. Officer Schwab did not recalibrate, but rather obtained two more samples which were valid.
On January 5, 1990, the machine was used again by Officer Campellone who tested a Mr. Gerald *593Kofoio. The results were a 0.155 and 0.132 percent. This 0.023 variance was explained by the officer on the assertion that the subject blew too softly on the machine. However, four tests taken after that were accurate.
On January 7, 1990, Officer Barry tested another subject, with results of a 0.245 and 0.223 percent. Poor samples were once again the basis upon which this deviation was explained away.
Officer Harran further utilized the machine on January 20, 1990 on a Mr. Sweeney. The test results of 0.239 and 0.262 percent were explained by the officer as being the result of a hard blow followed by a soft blow. However, no notation was made to this effect in the log book and the machine was not recalibrated.
Finally, on February 15, 1990, Officer Barry tested a Mr. Grow and obtained results of 0.272 and 0.228 percent, a variance of 0.044. Once again, no new calibration tests were performed.
Officer Schwab testified at trial that all of these deviations were satisfactorily explained and that therefore no reason existed to take the Intoximeter out of service. However, Dr. Richard Saferstein testified that the velocity of a breath sample should not yield a result greater than a 0.020 discrepancy.
Therefore, defendant argues that as a result of said deviations the machine should have been taken out of service until it was recalibrated or repaired as necessary. We agree.
As stated above, Pennsylvania law requires that a breath test operator must record two consecutive actual breath tests, the lower result of which may be used for prosecution. 67 Pa.C.S. §77.24(b)(1) to (2). The statute requires that the test results will be disregarded and the breathalyzer removed from *594service “[i]f the difference between the results of the two actual alcohol breath tests is 0.02 or more, for machines read to the second decimal place, or 0.020 or more for machines read to the third decimal place.” Id. at section 77.24(b)(2)(i).
The Commonwealth asserts that despite the statute, the test results should be admitted, and any evidence as to their accuracy should go to the weight of the testimony. The Commonwealth supports this assertion with several Pennsylvania cases. First, the Commonwealth states that a breathalyzer test should be admitted despite any malfunctions because a history of malfunctions goes to the weight of the test results rather than admissibility. Commonwealth v. Shiffler, 373 Pa. Super. 497, 503, 541 A.2d 780, 783 (1988). However, the Commonwealth’s reliance on this case is misplaced because of the facts. In Shiffler, the Superior Court held that malfunctions of the breathalyzer went only to reliability, as opposed to admissibility, because in that case the matter was not raised at a suppression hearing, but rather at trial. Id. at 502, 541 A.2d at 782. The threshold issue of admissibility was never raised at all. Therefore, Shiffler is not applicable.
The Commonwealth also relies on Commonwealth v. Sesler, 358 Pa. Super. 582, 518 A.2d 292 (1986) for the proposition that failure to calibrate should be a credibility issue at trial, not a ground for suppression. Id. at 586-7, 518 A.2d at 295. However, the regulations upon which Sesler was decided were different from the case at bar. In Sesler, the interim regulation did not require two consecutive breath tests nor did they mandate dismissal of the results based on a 0.02 deviance. Therefore, neither of these cases are in conflict with the statute upon which we rely.
*595Finally the Commonwealth relies on an opinion handed down by the Honorable Ward F. Clark in the Common Pleas Court of Bucks County in support of its contention that the breathalyzer did not malfunction. Commonwealth v. McMonagle, 58 Bucks L. Rep. 35 (1990). In that case the court found that no malfunction occurred thereby giving credence to a police officer’s testimony that an individual who tested about two weeks before the subject defendant “was too intoxicated to perform the test properly.” Id. at 36.
In the present case, there were at least seven deviations from the allowable 0.020 variance. Although an explanation that an improper breath sample was given might be credible in one or two instances, the fact that so many deviations occurred indicates that the machine should have been removed from service and fully checked and calibrated. Also, the fact that no notation was made as to the deviation in the test performed on Mr. Sweeney is a factor affecting the credibility of the Commonwealth’s case. Moreover, the test performed upon Mr. Grow which yielded a variance of 0.044 is the greatest indicia of machine malfunction. Despite the officer’s testimony that the subject performed an improper breath sample, this court was persuaded that serious questions about the accuracy of the machine were raised, and should have been a concern to the Bensalem Township Police Department under the circumstances. Therefore this court finds that the machine should have been taken out of service until it was recalibrated or repaired as required by the statute. Since the Commonwealth has offered no evidence that this was accomplished, we hold that this evidence should be suppressed under section 77.224 of Title 67.