OPINION
WOZNIAK, Judge.Young appeals from an order sustaining revocation of his driver’s license. We reverse.
FACTS
Young was arrested for driving while under the influence and agreed to take a breath test. Young was observed for the recommended 15-20 minutes before the test was administered. The Intoxilyzer performed an internal diagnostic procedure, which indicated the machine was functioning properly. The air blank and calibration standard tests yielded results within acceptable limits. Young provided two adequate breath samples. The first sample showed an alcohol concentration of .094, with a replicate reading of .097. The second breath sample gave a reading of .109, with a replicate of .110. The reported value was .09, and the breath correlation between the two samples was .88. The officer testified that the Bureau of Criminal Apprehension (BCA) training which he received indicated that the correlation should be 90% and recommended that if it was not, he should request a second test. The testing officer also testified that while the machine did not indicate that the first sample was deficient, he believed that the breath correlation was low because the first sample was the minimum which the machine could take and the second was an excellent sample.
Young took a second test and again provided two adequate samples. The reported value was .10 and the breath correlation was 99%. The officer, who has performed approximately 75 Intoxilyzer tests, saw nothing to make him believe the Intoxilyzer was not functioning properly.
Young testified that he gave an adequate breath sample both times he was asked to do so for the first Intoxilyzer test. He testified that the officer told him after the first test, “Looks like we have to take another one.” Young assumed that he had to take the second test.
Young’s driver’s license was revoked because the second test indicated an alcohol concentration of .10. Minn. Stat. § 169.-123, subd. 4 (1984). He petitioned for judicial review and a hearing was held. The testimony was taken before a referee, who recommended rescinding the revocation. The Commissioner moved for review of the recommended order and the trial court, after a hearing in which legal arguments were made, issued an order sustaining the revocation. Young appeals from the trial court’s order.
ISSUE
May a police officer request a second test, and may the Commissioner revoke a driver’s license based upon a result of .10 *214from the second test, because the first test, which indicated an alcohol concentration of .09, showed a breath correlation of .88 between the two breath samples?
ANALYSIS
Appellant challenges the use of the result of .10 from the second alcohol concentration test to revoke his license, contending that a DWI suspect may not be required to take more than one valid test, and that the first test, which revealed an alcohol concentration of .09, was valid and reliable.
“Any person who drives, operates, or is in physical control of a motor vehicle * * * consents * * * to a chemical test * * Minn.Stat. § 169.123, subd. 2(a) (Supp. 1985). If a test malfunctions, a driver may be required to submit to an alternative test. Gunderson v. Commissioner of Public Safety, 351 N.W.2d 6, 7 (Minn. 1984).
The driver here provided two adequate breath samples. See Minn.Stat. § 169.123, subd. 2b (1984). Despite the breath correlation of .88, all the factors indicated that the test was valid and reliable. Appellant was observed for 15-20 minutes prior to providing breath samples. Internal checks indicated the machine was functioning properly, and the air blank and calibration standard tests were within proper limits. See Abe v. Commissioner of Public Safety, 374 N.W.2d 788, 791 (Minn.Ct.App.1985). The result from the first test was an alcohol concentration of .09, which does not lead to revocation of the driver’s license.
The Commissioner argues that the low breath correlation was due to a poor breath sample, which led to an inaccurately low report of appellant’s alcohol concentration. See Wells v. Commissioner of Public Safety, 392 N.W.2d 721, 722-23 (Minn.Ct.App.1986). The machine, not the operator, determines whether the breath sample is adequate. Huber v. Commissioner of Public Safety, 382 N.W.2d 573, 575 (Minn.Ct.App.1986). The Commissioner also argues that this court has urged that the officers follow BCA recommendations. See Wells, 392 N.W.2d at 723. Under these circumstances, there was no authority for requiring a second test, where the first test was valid and reliable, and which indicated an alcohol concentration of .09.
In addition, the fact that appellant consented to the second test, which is not required by the statute, is not determinative. As his testimony indicated, as far as he knew, he had no choices other than taking the test or losing his license for one year. See Haugen v. Commissioner of Public Safety, 389 N.W.2d 222, 223 (Minn.Ct.App.1986) (where the driver was offered only a blood test rather than the choice between a blood or urine test as mandated by the legislature, no revocation should occur, even though the driver voluntarily took and failed the blood test).
Further, we cannot ignore the Commissioner’s inconsistency as it relates to correlation. The Commissioner’s position has been firm that, when the correlation is slightly below .90 but the Intoxilyzer is functioning properly, the revocation should stand if the alcohol concentration reading is .10 or above.
Our court has recently considered appeals by drivers arguing that a correlation below .90 does not provide a sufficient basis for revocation. We have consistently rejected this argument, confirmed the validity of the test, and upheld the Commissioner.1
Where does this lead us? It is simply this — when the alcohol concentration reading is above .10 but the correlation is slightly below .90, the Commissioner demands the right to give only one test and come into court with that test to seek a revocation. Should we decide now, as the Commissioner argues, that, if the correla*215tion is slightly below .90 and the alcohol concentration reading is below .10, the tester can arbitrarily request a second or third test, even when the breath samples were adequate, the Intoxilyzer is functioning properly, and the test was valid?
We are aware that the BCA guidelines recommend a correlation of at least .90. However, the Commissioner has consistently refused to consider that recommendation binding on him when the correlation is below .90, but the alcohol concentration reading is above .10.
Should the Commissioner or the legislature require that tests with a correlation of below .90 be discarded regardless of the alcohol concentration reading, at least a uniform guideline would then control.
That is not now the case, nor is it our prerogative as a court of review to insist that a particular guideline be implemented with the force of law. For the present, we simply hold that, if the Intoxilyzer is functioning properly and the breath sample is not deficient, then a driver is equally entitled to take advantage of a reading below .10 with a correlation of below .90, as the Commissioner likewise takes advantage of a low correlation but a high reading.
The trial court erred in determining that the test operator properly requested that appellant take a second test and in sustaining the revocation based upon the result from the second test.
In view of our decision, we need not reach two further issues Young raised. Saving further appeals on these issues, however, we choose to address them. Young argued that the trial court erred when it reviewed the referee’s recommendations without a transcript. See Minn.R. Civ.P. 53.05(1). He waived the requirement below and cannot now argue that the trial court erred when it did not consider a transcript. Young also asserted the trial court improperly rejected the referee’s findings. The trial court did not accept the referee’s recommended order because it disagreed as a matter of law, which it is entitled to do. See Appelgate v. Commissioner of Public Safety, 402 N.W.2d 106, 109 (Minn.1987).
DECISION
The order of the trial court sustaining the revocation is reversed.
Reversed.
NIERENGARTEN, J, dissents.
. E.g., Wells, 392 N.W.2d 721; Schwarzrock v. Commissioner of Public Safety, 388 N.W.2d 425 (Minn.Ct.App.1986); Daley v. Commissioner of Public Safety, 384 N.W.2d 536 (Minn.Ct.App. 1986); Hager v. Commissioner of Public Safety, 382 N.W.2d 907 (Minn.Ct.App.1986); Abe, 374 N.W.2d 788; Zern v. Commissioner of Public Safety, 371 N.W.2d 82 (Minn.Ct.App.1985).