delivered the Opinion of the Court.
¶1 Michael Duane Crawford (Crawford) was charged with five *482separate driving-related offenses, including felony driving under the influence (DUI). He pled guilty to three of these offenses-driving on a revoked license, failure to have motor vehicle insurance and failure to use a seat belt. The jury found him guilty of the other two-felony DUI and violating a red traffic signal. He appeals the felony DUI conviction. We affirm in part and reverse and remand for a new trial.
ISSUES
¶2 A restatement of the issues Crawford presents is:
1. Did the District Court erroneously deny Crawford’s motion in limine to preclude the use of the results of the preliminary breath test (PBT)?
2. Did the District Court erroneously allow Officer Malhiot to testify as an expert on the scientific basis for the horizontal gaze nystagmus (HGN) test?
FACTUAL AND PROCEDURAL BACKGROUND
¶3 Crawford was stopped by Officer Sorenson just before midnight on December 2, 1999, for failing to stop at a red flashing light. At Crawford’s trial, Sorenson described, without objection, his observations and actions that ultimately led him to conclude that Crawford was driving while under the influence of alcohol. He explained that after he initiated the traffic stop and asked Crawford for his license, registration and proof of insurance, he noticed Crawford’s fingers “were rather clumsy,” he smelled of alcohol, his eyes were “glazed and somewhat bloodshot,” and his “speech was somewhat slurred and thick tongued.” Also, according to Sorenson, Crawford admitted to having had two beers.
¶4 Sorenson asked Crawford to get out of his car and to perform field sobriety tests. Crawford told Sorenson that he had bad knees, so Sorenson did not ask Crawford to perform any balancing maneuvers. He asked Crawford to perform the HGN maneuver and Crawford showed four of the six HGN indicators of possible alcohol impairment. Sorenson then asked Crawford to recite the alphabet. Crawford could successfully recite the alphabet to the letter “G” only. Sorenson, with Crawford’s permission, administered a PBT that showed a breath alcohol concentration (BAC) of .153. These observations and tests provided probable cause for Sorenson to arrest Crawford for DUI and transport him to the Cascade County Detention Center, where Crawford refused further processing and testing, including specifically refusing to take an Intoxilizer test.
*483¶5 Crawford was charged by Information with felony DUI, among other charges. The District Court issued an Omnibus Hearing Memorandum and Order on March 8, 2000, and ordered that all pretrial motions be filed no later than four weeks before trial. On April 17, 2000, the first day of his trial, Crawford pled guilty to driving on a revoked license, failing to have motor vehicle insurance and failing to use a seat belt. Crawford also filed a written motion in limine to suppress evidence of his PBT results. The District Court denied the motion on the merits.. Additionally, the State made an oral motion to qualify Matt Malhiot as an expert witness for the purpose of laying scientific foundation testimony regarding HGN. The court granted the motion over Crawford’s objection.
¶6 On April 18, 2000, the jury returned a guilty verdict on both felony DUI and failing to stop for the red flashing traffic signal. On June 13, 2000, the District Court sentenced Crawford to the Department of Corrections for nine months for placement in a prerelease center, to be followed by three years probation, for felony DUI. Crawford received suspended sentences and fines for the misdemeanor charges. The District Court’s Judgment of Conviction and Sentencing Order was entered on August 15,2000. Crawford filed a timely appeal.
STANDARDS OF REVIEW
¶7 A district court’s ruling on a motion in limine is an evidentiary ruling and the court has broad discretion in determining whether evidence is relevant and admissible. As such, this Court will not overturn a district court’s determination absent an abuse of that discretion. Somont Oil Co., Inc. v. A & G Drilling, 2002 MT 141, 310 Mont. 221, 49 P.3d 598.
¶8 The determination of the qualification and competency of an expert witness rests within the sound discretion of the trial court and will not be disturbed absent a showing of abuse of discretion. State v. Clausell, 2001 MT 62, 305 Mont. 1, 22 P.3d 1111 (citation omitted).
DISCUSSION
¶9 The first issue we consider is whether the District Court erred in denying Crawford’s motion in limine to preclude the use of the results of the preliminary breath test (PBT).
¶10 In State v. Weldele, 2003 MT 117, 315 Mont. 452, 69 P.3d 1162, we analyzed and decided the precise issue presented by Crawford. We will not repeat our full analysis here but will reiterate some essential points and apply our holding.
*484¶11 In Weldele, we affirmed State v. Strizich (1997), 286 Mont. 1, 952 P.2d 1365, and held that despite the statutory language of § 61-8-404, MCA (1997), PBT results obtained under field conditions are not admissible at trial as evidence of BAC without a showing that the results are demonstrably accurate and reliable. We noted that since Strizich was decided, we have not been made aware of any scientific evidence that allays our concern that the PBT instrument remains inherently unreliable for the purpose of accurately quantifying BAC. We also opined in Weldele that because the PBT utilizes a scientific instrument, the result would likely be accorded greater weight by a jury or judge than non-scientific, subjective indications of intoxication, i.e., officer observations and performance on field sobriety tests. For these two reasons, we concluded that admission of PBT results into evidence to prove actual BAC was error. Weldele, ¶ 57.
¶12 We further observed in Weldele that proponents of the 1997 statutory changes sought to have PBT results admitted in conjunction with the arresting officer’s observations and interpretations of the defendant’s performance of field sobriety tests, but that the amended statute failed to incorporate this caveat. Weldele, ¶ 54.
¶13 In the case sub judice, as in Weldele, the State offered no scientific evidence establishing consistent reliability and accuracy of the PBT equipment used to test Crawford. However, relying exclusively on § 61-8-404, MCA (1997), the District Court admitted the PBT results at trial as evidence of Crawford’s intoxication. As we did in Weldele, we hold that this was an erroneous legal conclusion which led the District Court to abuse its discretion in denying Crawford’s motion in limine.
¶14 It is at this stage in our analysis that the facts in this case diverge from those in Weldele. After concluding that admission of Weldele’s PBT result was error, we concluded that it was harmless and nonprejudicial error. We reached this conclusion after conducting a “harmless error” analysis under State v. Van Kirk, 2001 MT 184, 306 Mont. 215, 32 P.3d 735. We first determined that admitting Weldele’s PBT result was “trial” error rather than “structural” error, and as a result was subject to a “harmless and prejudicial” analysis. Weldele, ¶ 61; Van Kirk, ¶¶ 38-40.
¶15 We next applied the “cumulative evidence test” developed in Van Kirk. The “cumulative evidence” test requires that the court have before it admissible evidence that proves the same fact as the tainted evidence was submitted to prove. Under this test, if the tainted evidence is cumulative and is presented for the purpose of proving an *485element of the charged crime, the State must prove that, qualitatively, the tainted evidence would not have contributed to the conviction. If, on the other hand, the tainted evidence is not cumulative and is presented to prove an element of the crime, such admission of evidence cannot be considered harmless and thus is reversible error. Van Kirk, ¶ 47.
¶16 In Weldele’s case, the jury was presented with various types of evidence admitted for the purpose of establishing the intoxication element of Weldele’s DUI charge. First, the arresting officer’s observations at the time of the traffic stop as well as the results of the field sobriety tests he conducted were admitted without challenge by Weldele. Weldele, ¶ 5. More importantly, Weldele’s PBT result and the result of his Intoxilizer test were presented to the jury. Weldele, ¶ 63. We concluded in Weldele’s case that because the more accurate Intoxilizer result was admitted, without challenge, to prove the intoxication element of the crime, admission of Weldele’s PBT result was merely cumulative and its presentation to the jury did not harm or prejudice Weldele. Weldele, ¶ 63. This is not the situation in the case before us.
¶17 As we noted above, Crawford refused to take the Intoxilizer test. Thus, the evidence presented to the jury to prove the intoxication element of Crawford’s DUI consisted of the arresting officer’s testimony and Crawford’s PBT result. While we believe that each of these types of evidence serves an important and necessary function, we do not believe they are qualitatively comparable.
¶18 As suggested above, there is a natural propensity among jurors to accord greater weight to objective scientific evidence than to subjective observations that are open to differing interpretations. If the scientific evidence is accurate and reliable, this is a desirable and fair result. However, if the scientific evidence in question has no demonstrable accuracy or reliability-as is the case with the PBT result in Weldele and here-then the error of its admission can have significant adverse consequences for the defendant. As we pointed out in Weldele, factors which affect the results of PBT tests administered in the field can produce readings that vary as much as 10%. Weldele, ¶ 53.
¶19 Returning to the case before us, the jury had before it not only the PBT results, but also the officer’s observations and the results of the field sobriety tests. Thus, arguably, the tainted evidence-the PBT result-was cumulative. However, we must still apply the analysis we adopted in Van Kirk, referenced here in ¶ 14: If the tainted evidence is cumulative and is presented for the purpose of proving an element *486of the charged crime, then the State must prove that, qualitatively, the tainted evidence would not have contributed to the conviction. Given the scientific nature of the PBT evidence and for the reasons set forth above, we conclude the State cannot meet this burden.
¶20 This is not to say that, without the PBT or Intoxilizer as admissible evidence, DUI defendants must go unpunished. Refusal to take either or both tests results in automatic suspension of the person’s driver’s license for a prescribed period of time. Section 61-8-402, MCA. Moreover, because our statute does not mandate that quantitative, numeric evidence of BAC is required to obtain a conviction, the State may still pursue prosecution for DUI by offering the defendant’s refusal to take the test(s) into evidence, as well as the field sobriety test results, the arresting officer’s observations and any available video tape of the field and detention center investigations. Section 61-8-404, MCA. See also State v. Holman (1990), 241 Mont. 238, 786 P. 2d 667. Moreover, as we indicated in Weldele, the “State may take the opportunity at any time to establish PBT reliability,” and if it succeeds in doing so, the results could then be admissible providing the evidence otherwise satisfied all other admissibility requirements. Weldele, ¶ 57.
¶21 The State points out that in Town of Columbus v. Harrington, 2001 MT 258, 307 Mont. 215, 36 P.3d 937, we found no error with the District Court admitting PBT test results. While the State is correct in this assertion, the explanation for that conclusion is simple: Harrington did not challenge the reliability or admissibility of the PAST/PBT results. The question presented in Harrington centered on the inconsistency between the newly revised statute and existing-and not revised-regulations. This Court, not presented with any claim of PAST/PBT unreliability, performed only the legal analysis for the issue presented: when a statute and a regulation are in conflict, which controls? We concluded that the statute controlled and, therefore, affirmed the District Court which had concluded the same. Here we are squarely presented with the question of whether PBT results should be admitted as evidence of BAC; thus our decision in Harrington is inapposite.
¶22 Because we conclude for the reasons set forth above that the State cannot demonstrate that the tainted evidence would not have contributed to Crawford’s conviction, the admission of the PBT results in Crawford’s trial was prejudicial error. We therefore reverse his conviction, and remand for a new trial. Upon retrial, the State may utilize all admissible evidence in making its case against Crawford, *487including Officer Malhiot’s testimony, as further addressed below. ¶23 The remaining issue presented is whether the District Court erred in allowing Officer Malhiot to testify as an expert on the scientific basis for the horizontal gaze nystagmus (HGN) test. Because we are remanding for a new trial, this issue must be resolved.
¶24 Our courts recognize that it is beneficial, and at times necessary, to have experts testify in order for judges and juries to better understand the evidence or facts presented in a case. “The test for admissibility of expert testimony is whether the matter is sufficiently beyond common experience that the opinion of the expert will assist the trier of fact to understand the evidence or to determine a fact in issue.” State v. Russette, 2002 MT 200, ¶ 11, 311 Mont. 188, ¶ 11, 53 P.3d 1256, ¶ 11 (citing Hulse v. State, Dept. of Justice, 1998 MT 108, ¶ 48, 289 Mont. 1, ¶ 48, 961 P.2d 75, ¶ 48). “Hence, in conducting an admissibility analysis, a district court must determine: (1) whether the subject matter of the testimony is one that requires expert testimony; and (2) whether the particular witness is qualified as an expert to give an opinion in the particular area on which he or she proposes to testify.” Russette, ¶ 11 (citing State v. Southern, 1999 MT 94, ¶ 49, 294 Mont. 225, ¶ 49, 980 P.2d 3, ¶ 49). The Russette Court in setting out this two-pronged rule simply restated Rule 702, M.R.Evid., which provides:
If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education may testify thereto in the form of an opinion or otherwise.
To establish a witness’s qualifications as an expert, Rule 702 implicitly requires a foundation showing that the expert has special training or education and adequate knowledge on which to base an opinion. Russette, ¶ 11.
¶25 As to the first prong of Rule 702 regarding the appropriateness of expert testimony pertaining to HGN tests, it was firmly established under Hulse that “the underlying scientific principle of the HGN test is still beyond the range of ordinary training or intelligence” and, as a result, it is an appropriate subject matter for expert testimony. Hulse, ¶ 69. Having met this first Rule 702 prong, the District Court next determined whether Malhiot was qualified as an expert under the requirements of Rule 702.
¶26 The State sought to have Officer Malhiot testify as an expert for the purpose of establishing the reliability of the HGN results. Such a *488foundation was required under Hulse. To establish that the test was scientifically reliable, Malhiot’s testimony would necessarily include a physiological explanation of the link between alcohol ingestion and nystagmus, this being the underlying scientific basis for the HGN test. Crawford, citing Hulse and Van Kirk, claimed that Malhiot, who was not a medical professional, was not qualified to act as an expert for this purpose.
¶27 In Hulse, the Court concluded that the arresting officer, having received 40 hours of training on various field sobriety tests, was qualified to administer and evaluate the test but was not qualified to explain the underlying scientific basis for the reliability of the HGN test. Hulse, ¶¶ 71-72. Similarly, in Van Kirk, the Court ruled that the arresting officer, who held a bachelor’s degree in medical technology and worked as a lab technician at Montana State Hospital, was qualified to administer the test and to evaluate the results, but did not have the special training or education to qualify him as an expert to explain the correlation between alcohol consumption and nystagmus. Van Kirk, ¶ 28. While we concluded that the arresting officers in both of these cases were not qualified to provide expert testimony as to the reliability of the HGN test, we did not hold that such an expert must be a medical professional, as argued by Crawford.
¶28 Unlike the officers in Hulse and Van Kirk, Malhiot presented a litany of credentials demonstrating his qualifications to give expert testimony regarding the reliability of the HGN test. Among others:
1. He holds a bachelor of science degree in criminal justice administration;
2. He has taken numerous courses in biology, anatomy, physiology, and forensic science;
3. He has developed a specialty in DUI and impaired driving enforcement, and had completed several courses in DUI enforcement through the Montana Law Enforcement Academy, the Forensic Science Division of the Montana Department of Justice, and the National Highway Traffic Safety Administration;
4. He completed both the Montana Highway Patrol’s field sobriety course and the instructor-level course through the National Highway Traffic Safety Administration;
5. He completed an instructor facilitator course with the National Highway Traffic Safety Administration as well as numerous other courses with the Division of Forensic Science;
6. He maintains currency in these areas by reading papers and studies devoted to DUI-related topics, including HGN and other *489field sobriety tests; and
7. He previously has been qualified as an expert for the purpose of laying scientific foundation for HGN in both justice courts in Cascade County and in proceedings before the Eight Judicial District Court Judge Kenneth Neill.
¶29 At trial, Malhiot explained the concept of nystagmus and discussed the correlation between alcohol in the human body and the phenomenon of nystagmus. He further testified to his knowledge of the types of nystagmus and the various causes, in addition to alcohol, which might lead to nystagmus.
¶30 We have previously held that a district court is “vested with great latitude in ruling on the admissibility of expert testimony.” Southern, ¶ 48 (citation omitted) (emphasis in original). We have also stated that “without a showing of an abuse of discretion, such a determination will not be disturbed.” State v. Detonancour, 2001 MT 213, ¶ 48, 306 Mont. 389, ¶ 48, 34 P.3d 487, ¶ 48 (quoting Cottrell v. Burlington Northern R. Co. (1993), 261 Mont. 296, 301, 863 P.2d 381, 384). In the case before us, given Malhiot’s extensive training and education, we cannot conclude that the District Court abused its discretion in allowing him to testify as an expert on HGN.
CONCLUSION
¶31 For the reasons stated in this Opinion, we reverse and remand this case to the District Court for retrial during which the State may pursue a DUI conviction utilizing admissible evidence. We affirm the District Court’s decision to allow Malhiot to testify as an expert.
JUSTICES NELSON, REGNIER and LEAPHART concur. JUSTICE RICE,concurring in part and dissenting in part.
¶32 I concur with the Court’s holding on the second issue, but respectfully dissent from the holding on Issue 1.1 would affirm.
¶33 The Court correctly notes in ¶ 13 that “the State offered no scientific evidence establishing consistent reliability and accuracy of the PBT equipment used to test Crawford.” However, the State was not challenged to do so.
¶34 On the day of trial, Crawford offered a straight-forward legal objection: that based upon Strizich and an administrative rule, PBT evidence was unreliable as a matter of law, and therefore, inadmissible. Crawford did not present evidence or argue that the PBT was unreliable based upon the underlying scientific basis for the test. ¶35 For the reasons set forth in my concurring opinion in Weldele, Crawford’s argument that PBT evidence was unreliable as a matter of *490law is incorrect. Strizich did not so hold, but further, the statutes upon which Strizich relied in determining that the evidence was statutorily inadmissible were revised by the Legislature in 1997. Thus, given the basis of Crawford’s objection, the District Court correctly denied it. The evidence is not unreliable as a matter of law.
¶36 However, that does not end the inquiry. Reliability remains a valid evidentiary issue. As Amicus Curiae, Montana Association of Criminal Defense Lawyers, correctly notes about the 1997 statutory revisions:
First, the Legislature did not mandate admission of the [PBT] results. The [statutory] provision does not direct that the results “must” or “shall” be admitted. Second, the Legislature did not declare the results were reliable scientific evidence. The Legislature merely removed the statutory impediment to admissibility of PBT results.
The statutory impediment to admissibility removed, PBT evidence may be offered, but remains subject to challenge on the grounds of its scientific reliability. When properly challenged on that ground, a trial court must consider the issue in the manner we set forth in Hulse. That is, as explained in my concurring opinion in Weldele, the trial court must “conduct a conventional Rule 702, M.R.Evid., analysis.” Hulse, ¶ 69.
¶37 However, no objection was made by Crawford to the underlying scientific basis of the PBT, and therefore, the State offered no evidence regarding its scientific reliability. Even if Crawford had objected to the evidence on the proper grounds, it would have been impossible for the State to demonstrate the test’s reliability, given the untimely nature of Crawford’s challenge. At any rate, the District Court did not rule on the issue, and on that basis, I would affirm.
CHIEF JUSTICE GRAY joins the concurring and dissenting opinion of JUSTICE RICE.