dissenting.
By its plain language, Georgia’s Dram Shop Act, OCGA § 51-1-40, shields servers of alcohol from liability for harm caused by an intoxicated person unless the intoxicated person was noticeably *366intoxicated at the time she was served alcohol,1 The majority ignores this legislatively-established standard and effectively replaces it with a new standard: whether the person who caused the harm was, in fact, intoxicated at the time she was served. Because the majority has changed the statutory duty from observing a person’s appearance to determining a person’s actual blood alcohol level, I respectfully dissent.
1. Although this Court has not previously considered whether expert testimony based on the tortfeasor’s blood alcohol content hours after she was served alcohol can create a jury issue regarding the tortfeasor’s appearance at the time she was served, we do not approach the matter without guidance. Courts in other states have considered this issue, but the majority fails to address these cases or the ways in which states with dram shop statutes similar to Georgia’s have handled the use of expert testimony. Instead, the majority says that Georgia’s law on summary judgment means a jury issue exists. Georgia law, however, requires entry of summary judgment when there is no genuine issue for the jury to resolve regarding a dis-positive part of the case.2 Here, Hulsey’s evidence, at most, addresses the average person’s reaction to alcohol, but as demonstrated by the weight of authority from other states, the average person’s reaction to alcohol does not create a genuine issue regarding whether the tortfeasor was noticeably intoxicated, which is a dispositive issue in this case.
Courts from other states have reached varying results in deciding the role of expert testimony in dram shop cases. Several courts have rejected attempts to use expert testimony to link the tortfeasor’s blood alcohol level later in the day to a point earlier in the day,3 whereas other courts have permitted expert testimony provided it focused on the tortfeasor’s reaction to alcohol,4 as opposed to the average person’s reaction.5 Only a few courts have not required expert testimony to explain the significance of a person’s blood alco*367hol level,6 and they have done so with little analysis.
Having considered the reasoning in the various cases, I would adopt a rule that permits plaintiffs to proceed to trial when the plaintiff’s expert focuses on the tortfeasor’s reaction to alcohol, but not allow a case to go forward based simply on generalized statements regarding the average person’s reaction. This balanced approach is consistent with the rationale of the vast majority of courts that have considered the issue.
Under this proposed test, when the non-moving party relies solely on the tortfeasor’s blood alcohol content to raise a genuine issue regarding whether the tortfeasor was noticeably intoxicated, an expert witness must explain what the tortfeasor’s blood alcohol level later in the day shows about the tortfeasor’s appearance earlier in the day. If an expert testifies that, given the tortfeasor’s blood alcohol level, the tortfeasor, or someone with similar characteristics, would have exhibited noticeable signs of intoxication when the defendant served her alcohol, then the jury must decide whether the tortfeasor was, in fact, noticeably intoxicated when the defendant served her. If, however, an expert cannot testify that a person with characteristics similar to the tortfeasor’s characteristics, including tolerance for alcohol, would have exhibited noticeable signs of intoxication when the defendant served her, then there is no way for a reasonable jury to find that the tortfeasor was noticeably intoxicated when served.
Requiring some connection between a subsequently obtained blood alcohol level and the tortfeasor’s appearance when served alcohol protects the legislature’s intent in passing the Dram Shop Act, while also ensuring plaintiffs will have their day in court if the evidence warrants it. Restricting the question of noticeable intoxication to eyewitness testimony would leave injured third parties without recourse when the eyewitnesses may potentially be biased in favor of the defendant, such as when all eyewitnesses are the defendant’s employees or the tortfeasor’s friends and co-workers. On the other hand, permitting a plaintiff to reach a jury based solely on a tortfeasor’s blood alcohol level hours after she was last served by the defendant or general assertions regarding alcohol’s effects on the average person would require servers to determine a person’s actual blood alcohol content regardless of how the person appears.
2. Turning to the facts in this case, the evidence is undisputed *368that Northside served Greene her last alcoholic beverage at 6:30 p.m. on January 2, 1992, and she left work around 8:00 p.m. Therefore, the critical issue on summary judgment was whether Greene was noticeably intoxicated at 6:30 p.m.
Northside moved for summary judgment and submitted sworn affidavits from eyewitnesses stating that Greene was not noticeably intoxicated at any time on January 2, 1992. To oppose Northside’s motion for summary judgment, Hulsey relied on Greene’s blood alcohol content as of 10:00 p.m. on the evening of the accident, as well as expert testimony based upon Greene’s blood alcohol level at 10:00 p.m. Greene’s blood alcohol content hours after she was served, without more, tells the factfinder nothing about her appearance at 6:30 p.m.7
Hulsey’s expert affidavit also does not create a genuine issue because the expert fails to link Greene’s blood alcohol level with how she would have appeared at 6:30 p.m. The affidavit certainly does not support the majority’s statement that “the expert testimony was that a woman with Greene’s blood alcohol level would have exhibited manifestations of intoxication.”8
Although the expert never states what Greene’s blood alcohol level would have been at 6:30 p.m,, a factfinder could extrapolate from other measurements in the affidavit that Greene’s blood alcohol content at 6:30 p.m. would have been between .15 and .18. The deficiency in the affidavit is that the expert does not connect this blood alcohol range to how Greene would have appeared, which is the exact point on which his testimony is needed to create a genuine issue.
In his affidavit, the expert states, “[a]t this level of intoxication [.18 to .21] there are a number of manifestations of intoxication. They will vary with the person’s history of drinking.” As the expert concedes, alcohol affects different people differently.9 Yet, nowhere does the expert testify that he has factored Greene’s drinking history into his analysis. Without having accounted for Greene’s drinking history, neither the expert nor a factfinder can have any idea whether Greene exhibited noticeable signs of intoxication when she was served alcohol.
Hulsey’s expert also fails to describe these “manifestations” or *369say that any of them would have been noticeable. For example, a manifestation such as blurred vision is not noticeable; on the other hand, sweating is observable, but not necessarily associated with intoxication. In contrast, bloodshot eyes, slurred words, or the inability to keep one’s balance may be signs that are noticeable and indicative of intoxication. Without knowing what manifestations existed or whether they would have been observable, a jury cannot determine whether any signs that existed would have been noticeable indicators of intoxication.
Additionally, the expert discusses women as a “general matter,” never stating whether a person with Greene’s characteristics would have exhibited any noticeable signs of intoxication that evening. The question at issue here is Greene’s appearance, not an average person’s appearance. Knowing how alcohol affects the average person or the reasonable person tells the factfinder nothing about whether a specific person’s intoxication was noticeable when the defendant served her alcohol.
For the same reasons, the statutory prohibition against driving with a blood alcohol level of .08 or greater sheds no light on whether a particular person appears intoxicated.10 OCGA § 40-6-391 represents a uniform level set by the legislature at which nobody may drive legally, regardless of actual ability to operate a motor vehicle safely at that blood alcohol level. In contrast, the legislatively-determined standard of noticeable intoxication is person specific. What matters is whether the particular person requesting service is in a “state of noticeable intoxication.”
Because none of the evidence in the record shows that Greene was noticeably intoxicated on January 2, 1992, the undisputed evidence in the record is that Greene was not noticeably intoxicated when Northside served her alcohol. Accordingly, the trial court correctly granted summary judgment to Northside, and the portion of the court of appeals’ opinion to the contrary should be reversed. Instead, however, a jury will decide whether Greene was noticeably intoxicated because the majority has rewritten the Dram Shop Act so that a person’s appearance is no longer the determinative fact.
Whether a person owns a bar or restaurant or merely is hosting a party at his or her home, they can no longer count on being protected by the Dram Shop Act when serving someone that appears sober because, under the majority’s holding, evidence of nothing more than the average person’s reaction to alcohol is now sufficient to defeat a motion for summary judgment and send the case to trial. If, as I believe, the General Assembly did not intend servers of alco*370hol to determine a person’s actual blood alcohol content before providing him or her with an alcoholic beverage, then the legislature should amend OCGA § 51-1-40 (b) to correct the error in the majority’s ruling.
Decided July 11, 2002 Reconsideration denied July 26, 2002. Savage, Turner, Pinson & Karsman, Brent J. Savage, C. Dorian Britt, H. Lehman Franklin, Jr., for appellant. Browning & Tanksley, Charles B. Tanksley, for appellee. Stokes & Murphy, Christopher T. Terrell, Winburn, Lewis, Barrow & Stolz, John J. Barrow, Irwin W. Stolz, Jr., amici curiae.I am authorized to state that Presiding Justice Sears and Justice Hines join in this dissent.
See also Birnbrey, Minsk & Minsk, LLC v. Yirga, 244 Ga. App. 726, 728 (535 SE2d 792) (2000); McElroy v. Cody, 210 Ga. App. 201, 202 (435 SE2d 618) (1993).
OCGA § 9-11-56 (c); Lau’s Corp. v. Haskins, 261 Ga. 491, 491 (405 SE2d 474) (1991); Western Heritage Ins. Co. v. Newcastle Auto Sales, 249 Ga. App. 262, 264 (547 SE2d 792) (2001).
Beaulieu v. The Aube Corp., 796 A2d 683, 691-693 (Me. 2002); Hollermann v. River Roost, No. C7-99-414, 1999 WL 639278 (Minn. Ct. App. Aug. 24, 1999); Romano v. Stanley, 684 NE2d 19, 23 (N.Y. 1997); Chiesa v. Delta M.R, Inc., 1997 WL 33343296, *3 n.2 (Mich. Ct. App. Oct. 7, 1997); Johnson v. Harris, 615 A2d 771, 776 (Pa. Super. Ct. 1992); Purchase v. Meyer, 737 P2d 661, 663-665 (Wash. 1987).
Bruce v. K.K.B., Inc., 52 SW3d 250, 255-256 (Tex. Ct. App. 2001); Godfrey v. Boston Old Colony, 763 So2d 12, 15 (La. Ct. App. 2000); Morrison v. Fleck, 697 NE2d 1064, 1070 (Ohio Ct. App. 1997); see also Douillard v. LMR, Inc., 740 NE2d 618, 622-623 (Mass. 2001).
Beaulieu, 796 A2d at 693; Hollermann, 1999 WL 639278; Johnson, 615 A2d at 776.
Stewart v. Ryan, 520 NW2d 39, 49 (N.D. 1994) (commonly understood appearance of person who consumed fifteen to twenty drinks creates jury question); Crocker v. Larson, 1990 WL 130087, at *2 (Term. Ct. App. Sept. 11, 1990) (blood alcohol of .23% an hour after accident, which occurred ten minutes after leaving bar created jury question); see also Vanderhoek v. Willy, 728 NE2d 213, 217 (Ind. Ct. App. 2001) (blood alcohol level coupled with appearance at scene of accident short time after being served).
See, e.g., Beaulieu, 796 A2d at 691; Douillard, 740 NE2d at 622-623; Finley v. Lake, 1998 WL 1988594, at *2 (Mich. Ct. App. Dec. 15, 1998); Rockwell v. Ullom, No. CV-313598, 1998 WL 563967, at *6 (Ohio Ct. App. Sept. 3, 1998); Romano, 684 NE2d at 23; see also Hollermann, 1999 WL 639278; Purchase, 737 P2d at 663-665.
Majority Op. at 364.
E.g., Beaulieu, 796 A2d at 693; Douillard, 740 NE2d at 622; Romano, 684 NE2d at 22; Purchase, 737 P2d at 664-665; see also Blackwell v. State, 259 Ga. 810, 811 (388 SE2d 515) (1990) (expert testified at trial that defendant had high tolerance for alcohol due to habitual drinking).
OCGA § 40-6-391 (a) (5).