In consolidated proceedings pursuant to Real Property Tax Law article 7 to review real property tax assessments for tax years 1991/1992 through 2000/2001, the petitioner appeals from an order and judgment (one paper) of the Supreme Court, Kings County (Pesce, J.), entered May 16, 2003, which, after a nonjury trial, denied the petitions and dismissed the proceedings.
Ordered that the order and judgment is modified, on the law and the facts, by deleting the provisions thereof denying the petitions and dismissing the proceedings concerning the tax as*667sessments for the tax years 1991/1992, 1992/1993, and 1997/ 1998, and substituting therefor provisions granting those petitions to the extent of finding over-assessments with respect to the tax years 1991/1992, 1992/1993, and 1997/1998, as follows:
“Tax Year Published Assessment Corrected Assessment
1991/92 $643,500 $537,750
1992/93 $643,500 $594,000
1997/98 $733,500 $702,000,”
and otherwise denying the petitions; as so modified, the order and judgment is affirmed, with costs to the petitioner.
Upon our review of the record, we are satisfied that the Supreme Court accorded the evidence the weight it was due (see W.T. Grant Co. v Srogi, 52 NY2d 496, 510, 511 [1981]; Matter of Erie Blvd. Hydropower, L.P. v Town of Ephratah Bd. of Assessors, 9 AD3d 540, 544 [2004]) with respect to its determination to adopt the recommendations of the respondents’ appraiser, except with respect to its determination to adopt the recommendations as to insurance expense. The evidence supported the use of the Rent Guidelines Board (hereinafter RGB) data for the years in which such data was available (1994-2000), and we otherwise apply the average RGB data available for the earlier tax years at issue. The actual amount of insurance expense adopted by the respondents’ appraiser was lower than the RGB data, and there was evidence that the owner of the subject property received a multiproperty discount on its insurance. Accordingly, we find that the average investor would rely on the RGB data.
Contrary to the petitioner’s contentions, the petitioner failed to establish that it was subject to unequal assessment as a result of the respondents’ use of the uniform New York City 45% class ratio rather than the lesser of the city ratio or New York State’s published class ratio. Pursuant to RPTL 720 (3) (b) (3) (c) and (d), in a challenge to a class ratio as being unequal, either party may submit as evidence “the latest applicable class ratio established for the roll containing the assessment under review” for special assessing units, such as New York City, and “the uniform percentage of value stated on the bill for the roll containing the assessment under review” for all assessing units. The statute expresses no preference between the state’s class ratio and the city’s uniform ratio. As a result, it was incumbent upon the petitioner to explain why the use of the city’s ratio would result in an unequal assessment. Instead, the petitioner’s expert adopted whichever ratio was less merely on the ground that it would result in a lower assessed value. In effect, this is *668an admission that the city’s ratio is acceptable. Accordingly, the Supreme Court correctly applied the city’s 45% ratio for all tax years at issue (see Matter of 49 Realty Co. v Commissioner of Fin., 15 AD3d 659 [2005] [decided herewith]).
Finally, although the petitioner is correct that the Supreme Court’s findings of fact must fall within the range of evidence in the record (see RPTL 720 [2]; Matter of Blue Hill Plaza Assoc. v Assessor of Town of Orangetown, 230 AD2d 846, 848 [1996]; Shore Haven Apts. No. 6 v Commissioner of Fin., 93 AD2d 233, 236 [1983]; cf. Matter of City of New York [A. & W. Realty Corp.], 1 NY2d 428, 432-433 [1956]; Bienenstock v State of New York, 287 AD2d 587, 588 [2001]; Zappavigna v State of New York, 186 AD2d 557, 560 [1992]), in the instant matter, it is clear from the Supreme Court’s determination that it intended to adopt the conclusions of the respondents’ appraiser. The respondents’ appraiser found an overassessment in three of the 10 years under review. The Supreme Court erred in finding that there was no overassessment, and it did not make new findings that the published assessments were accurate.
The following are this Court’s findings of fact, which adopt the calculations by the respondents’ appraiser, except with respect to the insurance expense, which is increased:
1991/92 1992/93 1993/94 1994/95 1995/96
Adopted Ins. Expense $14,537 $14,537 $14,537 $15,264 $15,900
Actual Ins. Expense $10,115 $ 7,281 $ 6,464 $ 7,241 $14,909
Ins. Exp. Difference $ 4,422 $ 7,256 $ 8,073 $ 8,023 $991
New Total Value $1,195,095 $1,320,280 $1,297,069 $1,392,029 $1,486,085
Rounded Value $1,195,000 $1,320,000 $1,295,000 $1,390,000 $1,485,000
City Ratio .45 .45 .45 .45 .45
Corrected Assessment $537,750 $594,000 $582,750 $625,500 $668,250
Published Assessment $643,500 $643,500 $579,150 $625,500 $625,500
Reduction $105,750 $ 49,500 $0 $0 $0
1996/97 1997/98 1998/99 1999/2000 2000/01
Adopted Ins. Expense $14,628 $14,628 $13,992 $13,356 $13,992
Actual Ins. Expense $10,097 $ 9,362 $10,195 $ 9,952 $10,000
Ins. Exp. Difference $ 4,531 $ 5,266 $ 3,797 $ 3,404 $ 3,992
New Total Value $1,568,391 $1,561,449 $1,753,474 $1,897,617 $1,870,476
Rounded Value $1,570,000 $1,560,000 $1,755,000 $1,900,000 $1,870,000
City Ratio .45 .45 .45 .45 .45
Corrected Assessment $704,250 $702,000 $789,000 $855,000 $841,500
Published Assessment $625,500 $733,500 $733,500 $760,500 $747,000
Reduction $0 $ 31,500$0 $0 $0
*669Thus, there was an overassessment in tax years 1991/1992, 1992/1993, and 1997/1998, and we modify the judgment accordingly. Schmidt, J.E, Santucci, Crane and Skelos, JJ, concur.