Minuteman International, Inc. (“Minuteman”) opposes the application by Royal Appliance Mfg. Co. (“Royal”) to register the mark MVP for domestic and industrial vacuum cleaners. The United States Patent and Trademark Office Trademark Trial and Appeal Board (“Board”) sustained the opposition based on its finding of likelihood of confusion. Minuteman Int’l, Inc. v. Royal Appliance Mfg. Co., Opposition No. 99,534, slip op. at 14 (TTAB March 3, 2000) (“Board Opinion”). For the reasons discussed below, we affirm the Board’s decision.
I
Minuteman manufactures commercial and industrial vacuum cleaners. It sells its vacuums on the commercial market (e.g., schools, hospitals, industry, offices, and contract cleaners) through distributors, janitorial supply dealers, and its own sales force. It advertises its vacuum cleaners primarily in trade magazines. In October of 1994, Minuteman introduced a new upright commercial vacuum that it identified with the mark MPV. Its subsequent application to register that mark was granted on November 21, 1995. See U.S. Trademark Registration No. 1,937,-213. The registration identifies the goods covered as “vacuum cleaners for commercial and industrial use .... ” The MPV vacuum sells for approximately $400.
Appellant Royal manufactures domestic and commercial vacuum cleaners. It produces several types of upright household vacuums under the mark MVP. MVP vacuum cleaners contain the most powerful motor currently allowed under industry standards. Applicant chose the mark MVP for these vacuums in order to communicate *966this idea of to aximum v acuum p ower. In most of its advertising, the phrase “Maximum Vacuum Power” appears below the acronym MVP.
Appellant sells its MVP vacuum cleaners to general consumers through retailers such as Wal-Mart. Royal also sells some MVPs through independent vacuum cleaner dealers and janitorial supply dealers. Appellant advertises its MVP vacuums in general consumer magazines, trade magazines, and on national television. The deluxe MVP sells for approximately $160.
In November of 1994, Royal filed an intent-to-use application, seeking to register the MVP mark for “electrical vacuum cleaners for both domestic and industrial use .... ” Minuteman filed an opposition under section 2(d) of the Trademark Act on November 3, 1995, alleging that MVP was confusingly similar to its own MPV mark. In response, Royal argued that the parties’ marks created different commercial impressions and that the channels of trade for their respective goods did not overlap.
The Board determined that there was no issue with respect to Minuteman’s priority, Board Opinion, slip op. at 9, and sustained Minuteman’s opposition on the basis of likelihood of confusion, id. at 14. It found that there was insufficient evidence to show that either MPV or MVP would be understood by relevant consumers as having any significance in connection with vacuum cleaners other than as a trademark. Id. at 11. As such, the Board analyzed the two marks as unpronouncable letter combinations and determined that their overall commercial impressions were substantially similar. Id. at 11-12. Looking at the descriptions in Royal’s application and Minuteman’s registration, the Board further found that Appellant’s goods were, in part, the same as Minuteman’s. Id. at 12.
Finally, the Board determined that the channels of trade for domestic and commercial vacuum cleaners overlapped to some extent, id. at 13, and that the channels for commercial and industrial vacuums were the same, id. at 14. Accordingly, the Board concluded that the contemporaneous use of the marks MPV and MVP on the same and related goods was likely to cause confusion. Id. Royal appealed. This court has jurisdiction pursuant to 28 U.S.C. § 1295(a)(4).
II
The United States Patent and Trademark Office (“PTO”) may refuse to register a trademark that so resembles a registered mark “as to be likely, when used on or in connection with the goods of the applicant, to cause confusion, or to cause mistake, or to deceive.” 15 U.S.C. § 1052(d) (1994). Likelihood of confusion is a question of law, based on underlying factual determinations. Recot, Inc. v. M.C. Becton, 214 F.3d 1322, 1326 (Fed.Cir.2000); Packard Press, Inc. v. Hewlett-Packard Co., 227 F.3d 1352, 1356 (Fed.Cir.2000). The Board and this court determine likelihood of confusion on a case-specific basis, aided by the factors set out in In re E.I. DuPont DeNemours & Co., 476 F.2d 1357, 1361 (CCPA 1973). Online Careline, Inc. v. America Online, Inc., 229 F.3d 1080, 1084 (Fed.Cir.2000); Packard Press, 227 F.3d at 1356; Recot, 214 F.3d at 1326.
The thirteen Dupont factors are: (1) the similarity or dissimilarity of the marks in their entireties as to appearance, sound, connotation and commercial impression; (2) the similarity or dissimilarity and nature of the goods described in the application or registration of the mark, or in connection with which a prior mark is in use; (3) the similarity or dissimilarity of established, likely-to-continue trade channels; (4) the conditions under which and the buyers to whom sales are made; (5) *967the fame of the prior mark; (6) the number and nature of similar marks in use on similar goods; (7) the nature and extent of any actual confusion; (8) the length of time during and the conditions under which there has been concurrent use without evidence of actual confusion; (9) the variety of goods on which a mark is or is not used; (10) the market interface between the applicant and the owner of a prior mark; (11) the extent to which the applicant has a right to exclude others from use of its mark on its goods; (12) the extent of potential confusion; and (13) any other established fact probative of the effect of use. DuPont, 476 F.2d at 1361.
The court is not obligated to analyze each DuPont factor in every case. Instead, it need only consider a factor when there is evidence of record on the issue. Id. See also Cunningham v. Laser Golf Corp., 222 F.3d 943, 946 (Fed.Cir.2000); Olde Tyme Foods, Inc. v. Roundy’s, Inc., 961 F.2d 200, 202 (Fed.Cir.1992). Additionally, the court “may focus its analysis on dispositive factors, such as similarity of the marks and relatedness of the goods.” Han Beauty, Inc. v. Alberto-Culver Co., 236 F.3d 1333, 1338 (Fed.Cir.2001) (citing In re Dixie Restaurants, Inc., 105 F.3d 1405, 1406-07 (Fed.Cir.1997)).
Finally, this court will uphold the Board’s factual findings on the DuPont factors unless they are unsupported by substantial evidence. On-line Careline, Inc., 229 F.3d 1080, 1085; Packard Press, 227 F.3d at 1356 (citing Dickinson v. Zurko, 527 U.S. 150, 165, 119 S.Ct. 1816, 144 L.Ed.2d 143 (1999)); In re Gartside, 203 F.3d 1305, 1315 (Fed.Cir.2000). The substantial evidence test requires the court to determine whether a reasonable person, looking at the record as a whole, could find that the evidence supported the Board’s conclusions. On-Line Careline, Inc., 229 F.3d at 1085-86. “[T]he possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency’s finding from being supported by substantial evidence.” In re Gartside, 203 F.3d at 1312 (quoting Consolo v. Fed. Mar. Comm’n, 383 U.S. 607, 620, 86 S.Ct. 1018, 16 L.Ed.2d 131 (1996)).
A
The first DuPont factor that we consider is the similarity or dissimilarity of the marks. DuPont, 476 F.2d at 1361. Royal argues that the evidence does not support the Board’s finding that the overall commercial impressions of MPV and MVP are substantially similar. According to Appellant, the Board’s determination that MVP is a common acronym for “most valuable player” precluded it from comparing MVP and MPV as two unpronouncable letter combinations that are inherently difficult to remember. Instead, Royal asserts, the Board should have analyzed MVP as a mark that evokes an attribute of Appellant’s goods, in keeping with V.I.P. Foods, Inc. v. Vulcan Pet, Inc., 210 USPQ 662 (N.D.Okla.1980), rev’d on other grounds, 675 F.2d 1106 (10th Cir.1982) and Gulf States Paper Corp. v. Crown Zellerbach Corp., 57 C.C.P.A. 720, 417 F.2d 795 (CCPA 1969).
We disagree. A determination of similarity or dissimilarity under DuPont requires an examination of the marks in their entirety; the Board, and this court, must consider all relevant factors pertaining to the marks’ appearance, sound, and connotation. See DuPont, 476 F.2d at 1361; Packard Press, 227 F.3d at 1357; Recot, 214 F.3d at 1329. Royal focuses exclusively upon the last factor, connotation, while ignoring the obvious similarities between the two marks’ appearance and sound. Even accepting arguendo Appellant’s assertion that MVP “suggests that the goods are representative of the upper echelon of vacuum cleaners,” we are still *968left with the clear similarity between MPV and MVP. As the Board noted, the two marks consist of the same three letters, and they both begin with an “M.” Board Opinion, slip op. at 10. Moreover, the last two letters in Royal’s mark, “VP,” are a transposition of the remaining letters in Minuteman’s mark, “PV,” and the marks sound alike when said aloud. Given this substantial evidence of similarity, we see no reason to disturb the Board’s finding on this issue.
Royal’s reliance on V.I.P. Foods and Gulf States Paper is misplaced. In V.I.P. Foods, the defendant accused the plaintiff of improperly seeking trademark protection rights in a common, descriptive term. 210 USPQ at 663-64. The district court rejected this argument on the ground that the mark was suggestive rather than descriptive or generic and was consequently entitled to protection without proof of secondary meaning. Id. at 666. V.I.P. Foods therefore presented a distinctly different question from the one at issue here and is inapposite. The same is true of Gulf States Paper. In that case, the Court of Customs and Patent Appeals found that the mark CZ was not confusingly similar to E-Z because the later “sounds like ‘easy1 and thus generates the idea of an attribute of the goods.” 417 F.2d at 797. The court determined that CZ, on the other hand, could “only represent initials of some sort, and one seeing or hearing the mark would probably presume that that is what it indicates.” Id. at 797-98. Because neither MPV nor MVP sounds like anything other than a combination of letters, Gulf States does not support Royal’s argument.
B
The next DuPont factor that we consider is the “similarity or dissimilarity and nature of the goods or services as described in an application or registration or in connection with which a prior mark is in use.” DuPont, 476 F.2d at 1361. As a preliminary matter, Royal argues that the Board abused its discretion when it refused to grant Appellant’s motion to modify the description of goods found in its application. Royal filed this motion more than a month after Minuteman’s extended testimony period had closed, seeking to change its identification of goods from “electrical vacuum cleaners for both domestic and industrial use” to “electrical vacuum cleaners for domestic use.” Board Opinion, slip op. at 6. According to Royal, this amendment would have accurately reflected the fact that MVP vacuum cleaners are not intended for industrial and commercial use.
The Board’s decision to deny Appellant’s motion is not reviewable by this court. As our predecessor court stated in Hercules Inc. v. Nat’l Patent Dev. Corp., “the question of whether the board abused its discretion in denying such a motion, filed after the testimony period, is a matter to be determined by the Commissioner of Patents and Trademarks not by this court.” 524 F.2d 1227, 1228 (CCPA 1975). See also Palisades Pageants, Inc. v. Miss Am. Pageant, 58 C.C.P.A. 1225, 442 F.2d 1385, 1388 (CCPA 1971) (“Whether or not the board abused its discretion in denying appellant’s motion in the face of this argument, however, was a matter to be determined, not by this court, but by the Commissioner of Patents, under well-settled principles of law.”); Martin v. Crown Zellerbach Corp., 57 C.C.P.A. 968, 422 F.2d 918, 919 n. 1 (CCPA 1970) (“[Ajppellant alleges error in the board’s refusal to let appellant, during the appeal period, amend the description of the goods. This, however, is not an appealable matter.”).
Having rejected Royal’s motion, the Board held that “applicant’s goods as identified in the application, ‘electrical vacuum cleaners for both domestic and industrial use,’ encompass opposer’s goods in its *969pleaded registration, ‘vacuum cleaners for commercial and industrial use.’ ” Minuteman Int’l, Inc., Opposition No. 99,534, slip op. at 12. The Board based this finding on its judicial notice of Webster’s II New Riverside University Dictionary (1984), defining “commercial” as “ [designating products, often unrefined, made and distributed in large quantities for industrial use.” Id. at 12 n. 7. The Board concluded that “commercial use” included “industrial use,” id., and accordingly held that “applicant’s goods are deemed to be, in part, the same as opposer’s goods.” Id. at 12.
We find no error in the Board’s conclusion. The law is clear that, in determining likelihood of confusion, the Board must look to the description of the goods contained in the opposer’s registration and the applicant’s application rather than to the goods’ actual use. See Packard Press, 227 F.3d at 1359 (“[T]he question of registrability of an applicant’s mark must be decided on the basis of the identification of goods or services set forth in the application, regardless of what the record may reveal as to the particular nature of the applicant’s goods .... ”); Cunningham v. Laser Golf Corp., 222 F.3d 943, 948 (Fed.Cir.2000) <“[T]he identification of goods/services statement in the registration, not the goods/services actually used by the registrant frames the issue.”). Here, the Board found that the descriptions contained in Minuteman’s registration and Royal’s application overlapped and consequently held that their goods were to a certain extent identical. Substantial evidence supports the Board’s finding on this issue.
C
Finally, we address the “similarity or dissimilarity of established, likely-to-continue trade channels.” DuPont, 476 F.2d at 1361. Royal argues that the Board erred when it held that the channels of trade for commercial and domestic vacuum cleaners overlap to the extent that Appellant sells some domestic MVP vacuums through janitorial supply stores. According to Royal, there is no evidence in the record to support the Board’s finding that it sells any domestic MVPs through commercial channels. As Appellant asserts, “the evidence shows that these [domestic MVP vacuum] cleaners were not resold through [commercial] channels, because the products are not marketed as commercial vacuum cleaners .... ”
Royal’s argument fundamentally misconstrues the nature of DuPont’s trade channel analysis. The question under this DuPont factor is not whether janitorial supply stores sold domestic MVPs os commercial vacuum cleaners, but whether they sold domestic MVPs and commercial vacuums, i.e. whether the domestic and commercial trade channels overlapped. Appellant relies upon the trial testimony of Richard Farone to support its assertion that janitorial supply stores “should not have resold the [vacuum] cleaners through commercial channels.” Such statements are irrelevant, however. The janitorial supply stores did not sell through commercial channels—they were the commercial channel. Thus we find that the Board’s conclusion is supported by substantial evidence.1
*970III
Royal finds an additional flaw in the Board’s analysis, arguing that the Board erred when it failed to consider evidence of no actual confusion between Appellant’s and Minuteman’s products. As this court has held, however, the relevant inquiry is likelihood of confusion, not actual confusion. Weiss Assoc., Inc. v. HRL Assoc., Inc., 902 F.2d 1546, 1549 (Fed.Cir.1990). It is unnecessary to show actual confusion in order to establish likelihood of confusion. Giant Food, Inc. v. Nation’s Foodservice, Inc., 710 F.2d 1565, 1571 (Fed.Cir.1983).
The decision of the Board is affirmed.
. Royal takes issue with the Board’s presumption that Royal's and Minuteman’s goods were sold in all of the normal channels of trade to all of the usual purchasers for goods of the type identified. It asserts that “[t]he Board should not have ‘presumed’ anything concerning channels of trade and relevant consumers in the absence of substantial evidence that MVP domestic cleaners were in fact being resold through trade channels that overlapped with Minuteman’s commercial trade channels.” Contrary to Appellant’s assertion, however, the law required the Board to make precisely this presumption in its DuPont analysis. It is well settled that absent restrictions in the application and registra*970tion, goods and services are presumed to travel in all appropriate trade channels to all potential purchasers of such goods. See Cunningham v. Laser Golf Corp., 222 F.3d 943, 948 (Fed.Cir.2000); Canadian Imperial Bank of Commerce v. Wells Fargo Bank, Nat’l Ass'n, 811 F.2d 1490, 1493 (Fed.Cir.1987); CBS, Inc. v. Morrow, 708 F.2d 1579 (Fed.Cir.1983).