We conclude that the method used by the county assessor to calculate the estimated market value of bare farmland for real estate tax purposes is acceptable and that the taxpayers’ land was not excessively valued. We affirm the tax court.
Relators Kenneth and Cindy Lamping claimed that the taxes assessed against their six parcels of farmland in Freeborn County for the years 1982 and 1983 (payable in 1983 and 1984) were based on excessive valuations and were unequally assessed in comparison with other land in the county. The tax court held that the method used in assessing the property was permissible; the court adopted the county’s valuations, except for making a downward adjustment in the valuation for the second year (1983) to account for a period of deflating values, and found no inequality with other lands. By certiorari, the taxpayers seek review before us. Relators do not now challenge the tax court’s finding that their property was treated equally as compared with other property in the county. They claim only that the valuations assigned their six parcels, even with the adjustments for the second year, are excessive.1
*171The County’s Assessment Method
To understand the issues raised by the taxpayers, we need first to describe the method of assessment used by the Freeborn County assessor. The assessor begins with a figure known as the crop equivalency rating, or CER.
Crop equivalency ratings were developed by the University of Minnesota, apparently in conjunction with the United States Department of Agriculture. The crop equivalency rating guide issued in 1984 explained that the ratings are intended to
reflect the relative net economic return per acre of soil when managed for cultivated crops, permanent pasture, or for forestry, whichever use is computed as giving the highest net return. An effort is made to express dollar equivalence in net return for the most commonly grown crops. * * * [A net return is derived] for all soils of interest and [the soils are then ranked] using 100 as equivalent to the highest rating.
R. Rust, L. Hanson & J. Anderson, Productivity Factors and Crop Equivalent Ratings for Soils of Minnesota 6 (1984); see also the original edition, R. Rust & L. Hanson, Crop Equivalent Rating Guide for Soils of Minnesota 7 (1975);. Thus the CER is a rating of the soil’s “potential” or “suitability” to “aid decisions about what land should be kept in agricultural use, what the intensity of soil management should be, what the cash unit or purchase value should be, and what might be considered as a basis for equalizing assessed valuation.” Id. at 9.
From 1968 to 1975, a survey team collected soil samples throughout Freeborn County, which were sent to the University of Minnesota for analysis. The results were then returned to the survey team which ultimately prepared a “soil survey map” of the county. CER’s were devised for each soil type.2 A chart for all types of soil in the county was prepared, setting forth the soil type, a description, and the crop equivalency rating.
A rating of 100 was assigned to the best soil in the county, “Nicollet clay loam,” soil type 130, with 0-3% slopes, and “Moland silt loam,” soil type 376-B, 2-6% slopes. Points were deducted for less desirable soil types. For instance, a soil that was sandy or “wet” would be assigned a number less than 100. A soil with flooding or drainage problems would be assigned a range of numbers for its CER. Thus, the CER for soil type 134 had a range from 5 to 80 CER. The survey team did not determine where a particular soil should fall within the assigned range. This determination was left to the county assessor.
The CER system assumes a moderately high level of farm management.3 Once soil *172has been typed and a CER assigned, the rating or range remains unchanged. Soil types are standardized throughout the state and CER values are the “same or very similar” county to county. The authors of the CER Guide state that the “selling price of bona fide sales of agricultural land has usually given an excellent correlation with the weighted average CER of the tracts involved.” R. Rust, L. Hanson & J. Anderson, Productivity Factors and Crop Equivalent Ratings at 11.
Freeborn County relied heavily on crop equivalency ratings in combination with comparable sales to determine market value of farmland for tax purposes. For the 1982 assessment, the county assessor first analyzed 55 arms-length farm sales in the county for the 24-month period from July 1979 to June 1981 and determined the average price per acre for the 55 sales to be $2,094 an acre. The assessor then determined the average CER for all the land in these sales, expressed as a percentage, to be .7395. Under the CER rating system, land having a CER of 100 is “perfect” land. Therefore, by dividing the average price per acre for the 55 Freeborn County sales ($2,094) by the average CER of the land (.7395), the assessor arrived at the value of perfect land with a rating of 100 to be $2,832 an acre.
To arrive at the value of a particular taxpayer’s parcel, the assessor first determined its average CER. He then multiplied the value of a perfect acre by this average CER to arrive at the market value of the taxpayer’s parcel. For example, if the average CER for a particular parcel was .75, the assessor multiplied the sale price of an acre of perfect land, which had already been calculated at $2,832, by .75, to arrive at an estimated market value of an acre of the taxpayer’s land of $2,124. Paul Knutson, the county assessor, relied exclusively on this use of comparative CER’s in his estimates of market value.
The assessor stated that the commissioner of revenue had directed that estimated market values should be at least 85% of actual market value as determined from comparative sales. Since land with a 100 CER would sell theoretically for $2,832 an acre during the period of July 1979 to June 1981, the assessor decided to value all lands in the county on the basis of land with a 100 CER having a value of $2,500 an acre.4 This theoretically resulted in valuing all lands on the basis of 88% of actual market value as determined by comparative sales. The assessor made no adjustments for time or financing terms in his calculations.
Clifford Davidson, the township assessor for the township in which the Lampings’ six parcels were located, visited all the farmland in his township, viewing the properties primarily from adjoining roads where available, and prepared field notes for each 40-acre tract. Using the soil map prepared by the survey team, he noted the number of acres of each soil type by 40-acre sections. From his CER chart he then assigned a CER number for each of the soil types. It will be recalled that soils with drainage problems were listed on the chart with a range of CER numbers. Nevertheless, Davidson testified that he always assigned this soil with the highest number in the range if the land was tilled.
The township assessor then sent his field notes to Paul Knutson, the county assessor, and Knutson calculated the estimated *173market value for each tract. Using the field notes, Knutson obtained the average CER for each parcel to be taxed. For each of the six Lamping parcels, Knutson multiplied the market value of one acre of “perfect” (100 CER) land by the average CER for the Lamping parcel in question. Thus the value of each Lamping parcel was a percentage of the value of a “perfect” parcel. For example, the value of Lamping parcel No. 4 for taxes assessed in 1982 was determined as follows:
$2,500.00 (88% of value per acre of 100 CER land)
X .6114 (average CER for parcel 4)
$1,528.50 taxable value per acre of parcel 4 land.
The tax court essentially endorsed this method. The court adopted the assessor’s estimated market value for the Lamping parcels for taxes assessed in 1982 and payable in 1983. The tax court, however, determined that farm values dropped from January 2, 1982, to January 2, 1983. Consequently, the tax court calculated the estimated market value for taxes payable in 1984 based on the adjusted values of six comparable sales proposed by relators’ expert, William Russell, rather than on the assessor’s sales analysis. This resulted in reducing the assessor’s 1983 estimated market values for the six Lamping parcels by an average of 14%.
The Taxpayers’ Contentions
Relators would apparently accept the county’s system of appraisal using the CER approach as a start for determining market value, but they claim the county then failed to make necessary adjustments to arrive at fair and accurate assessments. Relators presented two experts, both of whom used CER’s as a springboard. The first was William Russell, an agricultural consultant and appraiser, who prepared two reports, one for each assessment year, purportedly using three different approaches, namely, cost, market data, and income. Relators’ second expert was George Karvel, a professor in the College of Business at St. Cloud State University holding a chair in real estate, who devised a method by which the comparable sales used by the county could be adjusted for time and financing.
Broadly stated, relators contend that the county’s assessment method is defective because, citing Minn.Stat. § 273.12 (1984), it does not “consider and give due weight to every element and factor affecting market value.” In particular, taxpayers contend that the county’s method:
1) fails to adjust for time, i.e., for the fluctuation in farm values over a period of time;
2) fails to adjust for terms, e.g., fails to account for farmland sold on contract for deed where the purchase price, because of financing considerations, may be more than the cash sale price;
3) fails to adjust for the income potential of farmland; and
4) fails to adjust adequately for the physical characteristics of a particular tract, such as drainage, roads, and location.
Scope of Review
In a proceeding brought by a taxpayer challenging property valuations, there is a prima facie presumption that the assessor’s valuation is proper and the taxpayer has the burden of proving that the assessment is excessive. Matter of McCannel, 301 N.W.2d 910, 923 (Minn.1980). The tax court’s findings of fact must be sustained unless they are clearly erroneous and not reasonably supported by the evidence as a whole. McCannel, 301 N.W.2d at 923. Finally, since no one method alone can be used to estimate the value of all property, the assessor, subject to requirements of fairness and uniformity, may exercise his discretion in using one or more of the usual methods. De Luz Homes, Inc. v. County of San Diego, 45 Cal.2d 546, 564, 290 P.2d 544, 555 (1955), cited in Alstores Realty, Inc. v. State, 286 Minn. 343, 352, 176 N.W.2d 112, 118 (1970). With these propositions in mind, we now turn to a consideration of each of the factors that taxpayers claim were slighted by the tax court.
*174The county’s assessment method adequately considered all relevant factors.
1. Taxpayers do not find fault with the comparable sales used to derive the county’s market value estimates for their property or the county’s basic method of calculation. They contend, however, that the county erroneously assumes that the value of the comparables does not change over time. Thus, taxpayers point out that to determine the value of one acre of 100 CER land as of January 2, 1983, the county used sales for the period from January 1981 through September 1982, even though, as taxpayers claim, the value of farmland was falling during this period. We agree com-parables must be adjusted for time, nor do we understand that the county disagrees. Indeed, the notion of a comparable sale is that it can legitimately be compared to a theoretical sale of the parcel being appraised — and this would require sales reasonably related in time to the appraisal. The issue here is whether values, in fact, changed during the pertinent time periods.
The tax court found that the values of land from July 1979 to January 2, 1982, were relatively stable. This finding is not clearly erroneous and finds support in the evidence, including the testimony of Russell, relators’ expert. The assessor used the same land values for his January 2, 1983, assessment, and here the tax court found that farm values had deflated in the preceding year and adjusted the assessor’s values accordingly. In making the downward adjustment of approximately 14%, the tax court used the six comparable sales that relators’ appraiser, Russell, had used. We hold that the tax court’s findings in making the reduction for second year valuations are supported by the evidence and are not clearly erroneous.
2. Taxpayers next claim that the county erroneously assumed, in its use of compa-rables, that sales price always equals market value. Many farm sales, say taxpayers, are on contract for deed or use a purchase money mortgage or involve assumption of an existing mortgage with an interest rate below the prevailing rate. In these instances, say taxpayers, the sales price is inflated for favorable financing terms and, to that extent, does not reflect the value of the land. In other words, sales price and market value are not always synonymous. United National Corp. v. County of Hennepin, 299 N.W.2d 73, 76 n. 4 (Minn.1980).
Taxpayers, through the testimony of Dr. Karvel, proposed a method for making cash equivalency adjustments. These calculations, however, in large part were based on hypotheticals and unverified assumptions about normal financing terms and suffered from a lack of any actual, pertinent data. Dr. Karvel had no information about actual financing terms for the comparable sales that the county used. The assessor said he did not have this information, either, because the certificates of real estate value then being filed did not contain this information. The tax court demonstrated, however, that even excluding contract for deed sales from the county’s comparables, the cash price for 100 CER land was not greatly affected. While it may be that contract for deed financing will affect market value determinations at times, this was not established to be the case here. The tax court’s refusal to make a financing adjustment was not clearly erroneous.
3.Taxpayers next contend that the assessor failed to consider the income potential of the farmland. Taxpayers make two arguments: first, that the tax court erred in rejecting the income approach used by its expert, Russell; and, second, that the assessor erred in not using the results of his own rental survey.
Although agreeing with the Lamp-ings that income is a factor in valuing land, the tax court found Russell’s calculations unpersuasive. Russell admitted he did not use an ordinary income approach. Rather *175than use rental values, he made calculations on the amount of income the land could be expected to generate if it were used to produce corn. Russell admitted on questioning from the court that his income method was really a “questionable approach.” 5 The refusal to use Russell’s income approach was not clearly erroneous.
Nor do we think that the tax court can be faulted for not using the results of a rental survey conducted by the county assessor for the 1983 assessment. The assessor collected rental data in 1982 and 1983 by a survey sent to landlords and calculated a value of $2,200 an acre for 100 CER land for the 1983 assessment. It will be remembered that the assessor, under his market approach, had used a value for 100 CER land for the same year of $2,500 (88% of market value). See supra n. 4. Knut-son testified that he did not use this rental survey data in arriving at his valuations because it was not legislatively then required.6 While agreeing with taxpayers ■ that, income generation is a factor in determining land value, the tax court concluded that there was insufficient evidence in the record relating rentals to sales prices to permit a separate determination of value on the basis of capitalization of income. We agree. Moreover, a consideration of income is not entirely absent from the county’s market approach method. As the tax court noted, income is reflected in the sales prices of the comparables used. Moreover, to some extent, income potential is also reflected in the crop equivalency ratings. We hold, therefore, that the tax court did not err in refusing to adopt taxpayers’ approach to income potential.
4. Finally, taxpayers contend that the county assessor failed to adjust his market values for the unique characteristics of their farmland. Put another way,.taxpayers claim that the county erroneously assumed an exact linear relationship between the CER of a tract and its market value, ignoring such things as income potential, the tract’s location and shape, the actual drainage, roads, water supply, and any special problems.
Taxpayers illustrate their position by noting that Parcel 1 has severe drainage problems and that they have spent $20,000 to correct these problems, including installation of a lift station. They point out Parcel 6 has a steep slope which is hazardous to wheeled vehicles and difficult to farm. The CER number assigned to a parcel may, of course, take into consideration drainage and slope, and, indeed, the CER chart pro*176vides a range of numbers for land with drainage problems. Taxpayers point out, however, that if land with drainage problems "is nevertheless tilled and under cultivation, the practice of the county assessor is to assign the highest number in the CER range to that land. Taxpayers argue it is improper for the assessor to assume that simply because land is tilled it is adequately drained.
The tax court responded to this argument by pointing out that drainage and other parcel characteristics are still reflected in the market price of comparable sales and that the evidence did not indicate that the Lamping parcels differed from the comparables used any more than was indicated by the CER comparison. “We assume,” said the tax court, “that the sales were average in relationship to all these factors, and we find that the subject land was average, at least in all factors.” We conclude that this finding is not clearly erroneous. The township assessor, for example, testified that nearly every farm in the county has some sort of drainage problem. There was no evidence indicating that the Lamping tracts were below average for the factors involved.
SUMMARY
We are dealing here with an appraisal en masse of all bare land within a county, where it is important to arrive at fair, over-all valuations that avoid both assessments that are excessive and those that are unequal in comparison with all other land in the county. In this context, while all factors relevant to market value are to be considered, the degree to which these factors are singled out for attention may not be as detailed as in an individual fee appraisal.
All land is unique, and taxpayers can always point out how their property is “different.” The perplexing problem is how these differences are to be fairly accounted for in a mass appraisal for tax assessment purposes. No one approach is the only way it can be done. Here Freeborn County has chosen to use a market value approach which incorporates crop equivalency ratings. To a limited extent, the CER number assigned to a parcel reflects such factors as drainage, topography, and soil adaptibility, all of which bear on market value, and, indeed, studies have shown a direct correlation. Parcels are broken down into 40-acre tracts for CER numbers. Once a number assignment has been made, apparently it does not change. Land prices, however, do change. This change is accounted for by the use of comparable sales, which, in turn, yield an average market value of farmland in the county, a figure which, in turn, can be translated into the market value assigned to 1 acre of “perfect” 100 CER farmland. Consequently, the comparables must be sufficient to give a fair average market value and must be redetermined often enough to assure reflection of current conditions. Aberrations affecting an individual taxpayer’s property may still occur, especially when one is dealing in averages and with a local assessor’s judgment. Ordinarily, as part of the assessment process, these aberrations, when they occur, will also be called to the attention of the assessor or to the boards of review or equalization for correction. In this case, on this record, the findings of the tax court that relators’ land was not overvalued cannot be said to be clearly erroneous.
Affirmed.
. 1982 Valuations:
Assessor's EMV Taxpayers’ values Tax Court's values
Parcel 1 $ 121,300 $ 91,413 $ 121,300
2 216,500 140,590 198,500
3 178,900 149,508 178,900
1982 Valuations:
Assessor's EMV Taxpayers’ values Tax Court's values
Parcel 4 $ 97,400 $ 76,612 S 97,400
5 173,500 159,110 173,500
6 154,600 120,512 154,600
*1711983 Valuations:
Assessor's EMV Taxpayers' values Tax Court's values
Parcel 1 $ 117,400 $ 78,864 $ 100,300
2 182,000 118,635 142,400
3 188,100 131,801 164,900
4 97,400 66,404 84,900
5 173,500 132,185 151,200
6 152,500 104,105 133,200
NOTE: The assessor had made a mathematical mistake in computing the value of Parcel 2, which the tax court corrected. Also, the parties stipulated to the value of buildings on the land. Only the valuation of bare farmland is involved here.
.Carroll Carlson, a soil scientist with the United States Department of Agriculture, who headed the survey team in Freeborn County, testified that all of the characteristics of a particular soil are considered in assigning a soil type, "for instance, a drainage class, the texture, the depth of the profile, how far it's sand, gravel, how deep it is, water holding capacity, the erosion, the reaction, flooding frequency * * * [and] the lay of the land.” Topography is also considered. Thus the slope of land is indicated by a letter following the soil type number. "28B” is “Dickinson sandy loam” with slopes of 2-6%. "Natural drainage" but not artificial drainage is also considered.
.The following factors are inherent in the assumed management level:
1. Drainage on soils where necessary or protection from flooding;
2. Erosion control practices on soils where needed, including proper use of residues;
3. Use of limestone, sulfur, and fertilizer as indicated by soil tests;
4. Use of herbicides for weed control and insecticides as needed;
5. Use of adapted varieties [referring to the kind of seed] in populations [referring to the *172plant density] related to soil moisture and fertility supply;
6. Timeliness of all operations related to seeding, cultivation, weed and insect control, and harvesting; and
7. Harvest procedures that minimize losses. R. Rust, L. Hanson & J. Anderson, Productivity Factors and Crop Equivalent Ratings 4 (1984); see also the original edition, R. Rust & L. Hanson, Crop Equivalent Rating Guide for Soils of Minnesota 3 (1975).
. The calculations for the 1983 assessment were also based on the $2,500 figure, representing 88% óf the estimated market value of 100-CER land. The assessor analyzed 24 arms-length farm sales for the recommended 21-month period — January 1981 to September 1982 — and, finding the 1982 assessed market values for the property to be about 88% of the sales prices, decided to again use the 1982 value for 100 CER land.
. Russell testified he used three approaches— cost, market and income — to value the Lamping parcels. His "cost approach” was really a combination of a market approach for the bare , farmland and a cost approach for the buildings. His "market approach” used six comparables, adjusted for type of financing and time (deflation).
The tax court, while rejecting Russell’s income approach, did use Russell’s market (“cost”) approach to the extent the court determined that adjustments were necessary for time in the second assessment year.
. In 1981 the legislature directed that agricultural land be valued at the "lesser of its market value or the value which could be derived from its free market gross rental rate capitalized at a rate of 5.8 percent," effective for taxes levied in 1983 and payable in 1984. 1981 Minn.Laws, 1st Spec. Sess., ch. 1, art. 2, §§ 4, 25, codified at Minn.Stat. § 273.11, subd. 7 (Supp.1981).
In 1983, however, the legislature decided that subdivision 7 would not go into effect until the 1984 assessment year, for taxes payable in 1985, and that the "committees on taxes shall study the feasibility of assessing farmland on the basis of productivity and net earning capacity.” 1983 Minn.Laws, ch. 342, art. 2, §§ 27, 28.
Subdivision 7 of Minn.Stat. § 273.11 was repealed in 1984. 1984 Minn.Laws, ch. 502, art. 3, § 36. The 1984 legislature found that methods of determining the production value of farm property were not suitable as a basis for directly determining the value of individual parcels of farmland, and, ”[t]herefore, the legislature determines that market value should continue to be used as a basis for taxation but that that market value should be adjusted to reflect the production value of farm property.” 1984 Minn.Laws, ch. 502, art. 3, § 27.
The 1984 legislature also directed the commissioner of revenue to consider alternative methods of determining production value and to make a recommendation to it by January 1985. Ch. 502, art. 3, § 28.