Under the Americans with Disabilities Act, an employer cannot be held liable where the plaintiff presents "a direct threat to the health or safety of other individuals in the workplace." 42 U.S.C. § 12113(a)-(b). I agree with the majority that "[w]hether an employer has properly determined that a person poses a direct threat depends on `the objective reasonableness of [the employer's] actions.'" Maj. Op. at 270 (emphasis added) (quoting Bragdon v. Abbott, 524 U.S. 624, 650, 118 S.Ct. 2196, 141 L.Ed.2d 540 (1998)). But I part company with the majority's interpretation and application of the "objective reasonableness" standard.
Just because an employee disagrees with an employment decision - and just because a plaintiff's expert disagrees with the underlying medical judgment that led to that employment decision - does not make the employment decision "objectively *280unreasonable." To warrant trial, a plaintiff must present evidence that no reasonable person could agree with the employer's determination - evidence that it was objectively unreasonable to conclude that the plaintiff presented a direct threat to the health or safety of others. Michael Nall presented no such evidence here. All he says, in essence, is that he and his experts disagree with his employer and its experts. Accordingly, the district court was correct to render summary judgment for BNSF.
Under the majority's approach, however, a plaintiff is entitled to take a case to trial so long as he can identify some procedural irregularity in the process used by his employer to determine whether or not he presents a direct threat. I disagree. There is no basis for imposing liability under the ADA based on process concerns alone. There is liability only if the employer's determination of a direct threat is objectively unreasonable. And the majority has not identified any process irregularity that renders BNSF's medical conclusion objectively unreasonable. I respectfully dissent.
I.
The development of the "objectively reasonable" standard and its application in employment discrimination cases helps inform the way that we should apply the standard - and where the majority's analysis falters.
Generally, the "direct threat" defense balances the ADA's abhorrence of irrational bigotry toward the disabled with the reasonable safety concerns of industry and the need to give employers room to operate their businesses in good faith. See Bragdon, 524 U.S. at 649, 118 S.Ct. 2196 ("The ADA's direct threat provision stems from the recognition ... of the importance of prohibiting discrimination against individuals with disabilities while protecting others from significant health and safety risks.").
To maintain this balance, the Supreme Court first articulated the "objectively reasonable" standard under the public accommodations provisions of the ADA. See Bragdon, 524 U.S. at 649-50, 118 S.Ct. 2196. The Court held that the "existence, or nonexistence, of a [direct threat] must be determined from the standpoint of the person [allegedly violating the ADA], and the risk assessment must be based on medical or other objective evidence." Id. at 649, 118 S.Ct. 2196. The Court emphasized that courts should be concerned only about the "objective reasonableness" of the evidence. Id. at 650, 118 S.Ct. 2196.
The textual similarities between the public accommodations provisions and the employment provisions have led some of our sister courts to adopt the Bragdon standard for employment claims. See, e.g., Michael v. City of Troy Police Dep't, 808 F.3d 304, 307 (6th Cir.2015); Jarvis v. Potter, 500 F.3d 1113, 1122 (10th Cir.2007) ("We recognize that Bragdon was not an employment case.... But the Court explicitly pointed out that the ADA contains parallel language in its employment provisions and we see no reason not to apply Bragdon's analysis to employment cases.") (internal citations omitted).1
*281In a faithful application of the "objectively reasonable" standard, the question for juries is not whether an employer was "right" to find a direct threat to safety, but whether it was reasonable to do so. See, e.g., Jarvis, 500 F.3d at 1122 ("In evaluating an employer's direct-threat contention, the fact-finder does not independently assess whether it believes that the employee posed a direct threat."). The distinction between what is "right" and what is simply "reasonable" is critical, because the ADA does not forbid reasonable reliance on allegedly incorrect experts, but rather, irrational discrimination against the disabled. As a result, "[a] medical opinion may conflict with other medical opinions and yet be objectively reasonable." Michael, 808 F.3d at 307 (citing Bragdon, 524 U.S. at 650, 118 S.Ct. 2196).
So it is not enough that experts disagree over whether or not Nall presents a direct threat to the health or safety of others. Rather, there must be evidence that it was objectively unreasonable for BNSF to reach that conclusion.
II.
BNSF reasonably relied on the evidence available to conclude that Nall was a "direct threat." The record in this case shows that BNSF employees had sincere (and serious) concerns about Nall's ability to do his job - concerns that were far from dispelled by Nall's independent evaluations. Moreover, BNSF's field test confirmed its employees' concerns, and demonstrated that Nall could not do his job safely. Nall responds, in essence, that he and his doctors disagree.
Nall and BNSF agree about his core job responsibilities: operating switches, coupling air hoses, making quick hand and leg movements, performing training and equipment inspections, climbing on and off equipment, and maintaining good balance and a steady gait. Nall agrees that conductor and switchman are physically demanding positions for which safety is paramount. Furthermore, Nall admits that, on top of the express, written requirements of his job, it is important for him to be able to react quickly and to maintain his balance.
Nall's co-workers were the first to notice that something was wrong. They told BNSF that Nall was taking a long time to repeat information back to the dispatcher, and that he was also having difficulty walking along the track. BNSF memorialized these concerns in its original "Notice of Leave": employees "observed [Nall] having difficulty getting on and off engines as a result of stumbling as well as displaying a lack of concentration when receiving mandatory directives from dispatchers." Nall's response is essentially that he disagrees with his co-workers' characterizations.
After initially being placed on leave, Nall consulted four different doctors. They all reached the bottom line conclusion that Nall could perform his job. But three of them also confirmed the very concerns that gave BNSF pause. First, Dr. Nishith Majmundar observed that Nall had cogwheeling rigidity (or stiffness) and problems with balance (for which Nall blamed his shoes), and accordingly recommended that Nall increase the dosage of his medication. Second, Clinical Neuropsychologist Stephen Lippold noted some short-term verbal memory problems. Third, Licensed Occupational Therapist Mary Karasek noted that Nall had some problems walking on an uneven surface, a "mild, shuffling" gait, difficulty walking on a treadmill, and a delayed reaction time. Her conclusion was that Nall was able to do what his job *282required, but that his balance could be a concern. Only Doctor Joseph Jankovic was able to give Nall an unequivocal statement of support. He concluded that Nall was able to do his job, and that his "balance and gait has never been an issue."
Not surprisingly, then, these evaluations did not assuage BNSF. Despite their positive bottom-line conclusions, they also suggested that the concerns of Nall's co-workers might be valid. So BNSF asked Nall to take part in a field test that required him to perform various tasks under supervision. Though he successfully completed the tasks, one employee described Nall's field test as "[p]robably one of the most blatant `there is no way this guy is safe' field tests" she had ever seen. Findings from the field test highlight that Nall was confused while interpreting car numbers and the "switch list"; he had difficulty holding on to the equipment towards the end of the ride; he walked between moving equipment; he gave the wrong signal to the engineer; and he exhibited other minor physical problems that concerned observers. After the field test, BNSF told Nall he was still not safe to return to work.
Nall agrees that the field test required him to do things he would normally have to do for his job, but he disagrees with the conclusions of the field test. He disputes that he broke any rules. He also disputes some of the characterizations of his abilities. For example, he claims he did not have trouble with his balance. According to Nall, any concerns about his balance are just "what [the physical therapist] feels like, but [Nall] didn't have a problem with his balance." Nall "personally think[s he] did great on the test." But he does not claim that the therapist who wrote the evaluation had any reason to make up the observations. The closest he comes to disputing the legitimacy of the field test is saying that BNSF was "looking for something wrong" with his performance, and that it was focusing on the bad rather than the good.
III.
At most, the evidence cited in Nall's favor shows that his experts disagree with BNSF's experts on their bottom-line conclusions. But it was entirely reasonable for BNSF to give more weight to its own evaluation. After all, the things that BNSF evaluated in the railyard field test are not the kinds of things that doctors could test in their offices.
The evidence marshaled by Nall and the majority does not reach the core question of whether or not BNSF was objectively reasonable in finding a direct threat to health or safety. At most, it establishes merely a good faith dispute between experts after an orderly process: Long after BNSF first learned about Nall's original diagnosis, his co-workers brought safety concerns to the company's attention. Nall was then evaluated by independent doctors who confirmed some of those very same concerns. BNSF then conducted a field test to observe whether those concerns would manifest themselves in his job performance. And because they did, BNSF reasonably concluded that he was unsafe to return to work.
The majority addresses the record evidence at too high a level of generality. For example, the majority says that the reports from Nall's doctors "concluded that Nall could safely perform the tasks of a trainman listed on the medical status form that BNSF originally provided to Nall." Maj. Op. at 272. As noted, however, three out of Nall's four doctors confirm concerns about his balance and physical ability.
Under the objective reasonableness standard, we do not simply divide the evidence into "generally positive" and "generally *283negative" categories, announce the existence of a fact dispute, and then proceed to trial. Rather, we must separately determine whether any evidence exists that the employer was objectively unreasonable in relying on the negative evidence. Otherwise, all it would take is disagreement by a single doctor to foreclose summary judgment - even if the employer reasonably relied on expert medical opinion.
The evidence cited by Nall and the majority creates no issue of fact concerning whether BNSF behaved reasonably and relied on reasonable medical opinion. First, the majority cites medical reports from Nall's experts concluding that Nall could do his job safely. But these reports do not question the integrity of BNSF's evaluation. To the contrary, they confirm some of BNSF's concerns about Nall's physical and mental ability. Second, the majority asserts that Nall completed all of the tasks in his first field test. While true as far as it goes, it was the manner in which Nall completed the tasks that led to BNSF's decision. Third, the majority contends that we have previously found a disputed field test sufficient to preclude summary judgment. But in EEOC v. E.I. Du Pont de Nemours & Co., we held only that an employee who fails a field test at first, but then later demonstrates that she is capable of performing her job duties, is entitled to present that dispute to the jury. See 480 F.3d 724, 728 (5th Cir.2007) ("After the [1999 field test] confirmed [the employee's] walking impairment, [the employer's] physicians concluded that she should be medically restricted from walking anywhere at the plant.... [The employee's] attempt to get her job back was rebuffed by [the employer], even though she demonstrated in 2003 that she could walk an evacuation route without assistance."). That is not this case. To the contrary, the evidence here if anything indicates that Nall's condition has worsened, not improved. Finally, the majority notes that Nall disputes some of BNSF's evaluation and his co-workers' evaluations. But mere disagreement does not create an issue of fact - because it does not establish the unreasonableness of BNSF's reliance on those evaluations.
None of this evidence purports to show that BNSF relied on an objectively unreasonable opinion. At most, the evidence would simply allow the jury to reach a different conclusion about the competing experts. Cf. Michael, 808 F.3d at 307 ("An employer's determination that a person cannot safely perform his job functions is objectively reasonable when the employer relies upon a medical opinion that is itself objectively reasonable.") (internal citation omitted). To proceed to trial under these circumstances would replace the objective reasonableness standard with a de novo review of the employer's experts.
IV.
According to the majority, "the question on appeal is not whether it was reasonable for BNSF to conclude that Parkinson's disease symptoms prevented Nall from safely performing his duties; the question is whether BNSF came to that conclusion via a reasonable process that was not, as Nall alleges, manipulated midstream to achieve BNSF's desired result of disqualifying him." Maj. Op. at 271. I disagree with the majority's process concerns for two reasons.
To begin with, I see no basis for a process-based liability theory under the ADA. To my mind, Nall can prevail only if BNSF lacks a reasonable medical basis for concluding that he presents a direct threat to health or safety - and not based on process concerns alone. After all, an employee is a direct threat if he poses a "significant risk to the health or safety of others that cannot be eliminated by reasonable accommodation."
*28442 U.S.C. § 12111(3). And as the majority acknowledges, we ask only whether it was reasonable for the employer to conclude that its employee was a direct threat - not whether it was correct.
So an employer cannot be held liable if it reaches a reasonable medical judgment that an employee presents a direct threat to the health or safety of others. The ADA does not impose liability based on perceived procedural irregularities alone. To quote the Sixth Circuit: "An employer's determination that a person cannot safely perform his job functions is objectively reasonable when the employer relies upon a medical opinion that is itself objectively reasonable." Michael, 808 F.3d at 307 (emphasis added) (collecting cases). See also, e.g., Bragdon, 524 U.S. at 650, 118 S.Ct. 2196 ("[C]ourts should assess the objective reasonableness of the views of health care professionals."). An employer can only be held liable based on process concerns if the procedural irregularities render the employer's medical conclusion unreasonable.
Second, the majority has not identified any process concerns that should call into question the reasonableness of BNSF's medical conclusions. The majority primarily faults BNSF for what it characterizes as shifting job requirements and disqualification criteria - specifically, the addition of quick reaction times and good balance as essential job criteria. But Nall himself conceded that his job required him to have quick reaction times and good balance:
Q. Do you agree it's important when you're performing the conductor job or the switchman job or another trainman job, it's important to be able to react quickly if necessary?
[Nall]. Yes, sir.
Q. Okay. And do you agree it's important to be able to maintain balance and not slip or fall when you're working those positions?
A. Yes, sir.
And it was precisely Nall's reflexes and his lack of balance that concerned BNSF. In contrast, in Riel v. Electronic Data Systems Corp., the parties actually disagreed about essential job functions, thereby creating an issue of material fact for a jury to resolve. See 99 F.3d 678, 682-83 (5th Cir. 1996).
Moreover, as previously noted, among Nall's own four doctors, three of them expressed concern with his abilities. And the field test further confirmed all of BNSF's fears. In sum, the process employed by BNSF yielded reasonable medical judgments on which BNSF reasonably relied.
* * *
Although the majority opinion appears to state the applicable legal standard correctly, it is the majority's application of that standard - and not just its mere recitation - that will form circuit precedent. The majority has set this circuit on a road that is contrary to both the vision established by Congress and the very rule that it purports to adopt. I respectfully dissent.
Compare 42 U.S.C. § 12113(b) (providing an employer may have "a requirement that an individual shall not pose a direct threat to the health or safety of other individuals in the workplace") and 42 U.S.C. § 12111(3) ("The term `direct threat' means a significant risk to the health or safety of others that cannot be eliminated by reasonable accommodation.") (employment), with 42 U.S.C. § 12182(b)(3) ("Nothing in this subchapter shall [apply] where [an] individual poses a direct threat to the health or safety of others. The term `direct threat' means a significant risk to the health or safety of others that cannot be eliminated by a modification of policies, practices, or procedures or by the provision of auxiliary aids or services.") (public accommodations).