This is a landowners’ appeal from a judgment of a committee of three state trial referees (trial court), who, exercising the powers of the Superior Court, reassessed the damages for the taking by eminent domain of the plaintiffs’ real property in the town of Westport. The plaintiffs claim that the trial court should have: (1) found that a residential subdivision was the highest and best use of the undeveloped land; and (2) employed the appraisal methodology used by both parties. We affirm.
The undisputed facts are as follows. The subject property before condemnation consisted of 32.01 acres and is located at the intersection of Post Road East and Compo Road North in Westport. In January, 1988, the defendant, the town of Westport, filed a certificate of taking on the Westport land records whereby it took title to an undeveloped 29.495 acre portion of the plain*404tiffs’ property. This partial taking left the plaintiffs with 2.515 acres of residential property containing a single residential dwelling. Of the taken portion, 27.73 acres were located in a Residence A zone and the remaining 1.71 acres were within a General Business District zone. As compensation, the defendant deposited $8,375,000 with the Superior Court, which sum the plaintiffs have received.
The plaintiffs appealed to the Superior Court, claiming that the defendant had undervalued their property.1 The plaintiffs presented at trial the testimony of two appraisers, Ronald Glendinning and Donald Kennedy, concerning the fair market value of the property as of the taking date. Glendinning appraised the property at $14,665,000, and Kennedy valued it at $15,600,000. The town similarly offered the testimony of two appraisers, George Shawah and Kenneth Carvell, who valued the land at $9,250,000 and $9,009,000 respectively. The trial court, while agreeing with the plaintiffs that the land had been undervalued, did not adopt the methods used by any of the appraisers, and independently determined the fair market value of the property to be $9,425,000. The trial court ordered the defendant to pay the plaintiffs the additional sum of $1,050,000, with interest from the date of taking, along with $12,000 in appraisal fees and $1750 for expert fees. Thereafter, the plaintiffs appealed to the Appellate Court and we transferred the matter to ourselves pursuant to Practice Book § 4023.
I
The plaintiffs first claim that by rejecting their appraisal method, i.e., the lot method, in which fair market value is determined by the value of finished lots *405less the capital costs of all the improvements that would be necessary to put them in that condition, the trial court inferentially rejected a proposed residential subdivisión as the highest and best use of the property. The plaintiffs argue that the factual predicate used by the trial court to determine the fair market value was flawed because of the trial court’s failure to adopt such use as the highest and best use. Because we conclude that the trial court never rejected a subdivision as the highest and best use, we disagree.
“ ‘The owner of land taken by condemnation is entitled to be paid just compensation. . . .’ ” Minicucci v. Commissioner of Transportation, 211 Conn. 382, 384, 559 A.2d 216 (1989); Conn. Const., art. I, § 11. The amount that constitutes just compensation is the market value of the condemned property when put to its highest and best use at the time of the taking. Minicucci v. Commissioner of Transportation, supra; Cappiello v. Commissioner of Transportation, 203 Conn. 675, 681, 525 A.2d 1348 (1987); Budney v. Ives, 156 Conn. 83, 88, 239 A.2d 482 (1968). In determining market value, “ ‘it is proper to consider all those elements which an owner or a prospective purchaser could reasonably urge as affecting the fair price of the land . . . .’ Budney v. Ives, supra.” Greene v. Burns, 221 Conn. 736, 745, 607 A.2d 402 (1992). “ ‘The “fair market value” is the price that a willing buyer would pay a willing seller based on the highest and best possible use of the land assuming, of course, that a market exists for such optimum use.’ Mazzola v. Commissioner, 175 Conn. 576, 581-82, 402 A.2d 786 (1978).” Minicucci v. Commissioner of Transportation, supra, 384. “The ‘highest and best use’ concept, chiefly employed as a starting point in estimating the value of real estate by appraisers, has to do with the use which will most likely produce the highest market value, greatest financial return, or the most profit from the *406use of a particular piece of real estate.” State National Bank v. Planning & Zoning Commission, 156 Conn. 99, 101, 239 A.2d 528 (1968). “In determining its highest and best use the trial referee must consider whether there was a reasonable probability that in the reasonably near future the subject property will be subdivided.” Minicucci v. Commissioner of Transportation, supra, 385.
While we agree that all the appraisers operated under the assumption that a residential subdivision was the highest and best use of the property and that they all used the lot method to determine its fair market value, we do not agree that the trial court, in rejecting the lot method, failed to consider the highest and best use of the property. The trial court rejected as too speculative, given the evidence presented, the lot method of determining the fair market value of the proposed residential subdivision. The trial court instead determined that, under the facts of this case, the proper method for arriving at the fair market value was to evaluate the land in its condition at the time of the taking, i.e., raw and undeveloped, “ ‘with consideration given to any increment or enhancement in value due to the property’s present adaptability to subdivision development.’ (Emphasis added.) 4 P. Nichols, Eminent Domain (3d Ed.) § 12.3142 [1] [a], pp. 12-335-356.” Minicucci v. Commissioner of Transportation, 211 Conn. 382, 385, 559 A.2d 216 (1989). In order to value the land, the trial court had at its disposal evidence of sales of comparable land in addition to the raw data used by the appraisers in applying the lot method. Because the trial court specifically stated that it had considered the property’s use as a subdivision in valuing the land, we conclude that it did not reject such proposed use as the highest and best use of the land.
II
The plaintiffs next claim that the trial court improperly refused to establish the fair market value of the *407land by appraising the undeveloped property based upon its potential for subdivision into residential building lots, i.e., the lot method. The plaintiffs argue that the trial court: (1) improperly concluded that the lot method was barred as a matter of law; (2) improperly concluded that the lot method required the landowner to have taken actual steps to subdivide his property; (3) abused its discretion in rejecting the unanimous opinion of the four appraisers that the lot method was the best way to determine the fair market value of the land for a residential subdivision; and (4) improperly concluded that the lot method was not applicable given the facts of this case. We address each claim seriatim.
At trial, the plaintiffs offered evidence tending to show that the residential property could be subdivided into forty-six one-half acre lots with a value of between $227,000 and $250,000 per lot. By utilizing the lot method, i.e., multiplying the value per lot by the number of lots, the plaintiffs’ experts appraised the residential portion of the property at values within the range of approximately $10,400,000 to $11,500,000.
A
Initially, the plaintiffs claim that the trial court improperly interpreted Minicucci to stand for the proposition that the lot method valuation is invalid as a matter of law when dealing with undeveloped land. We disagree.
The trial court rejected the lot method because it concluded that the plaintiffs’ evidence as to the value of hypothetical completed subdivision lots was too speculative, not because it was improper as a matter of law. In rejecting the lot method, the trial court concluded that “[i]n the instant case there are a multitude of uncertainties present, as elucidated previously, that make the ‘lot method’ approach to valuation improper.” Further, we do not agree with the proposition that “the *408lot method of appraisal should never be admitted in condemnation cases involving unimproved raw land. . . . We think the better view ... is that a lot method appraisal can be admitted in appropriate cases if the proponent offers credible evidence of the costs of subdivision—e.g., the expense of clearing and improving the land, surveying and dividing it into lots, advertising and selling, holding it, and paying taxes and interest until all lots are sold. . . . The potential value of land if subdivided could well be considered by a willing buyer and a willing seller where subdivision is a reasonable possibility and the costs of subdivision are not speculative or uncertain.” United States v. 47.3096 Acres, Etc., In Oxford Township, 583 F.2d 270, 271-72 (6th Cir. 1978). In the instant case, the trial court simply concluded that the plaintiffs’ evidence was too speculative and uncertain to allow the use of the lot method of valuation.
B
The plaintiffs next claim that the trial court’s reliance on Minicucci for the proposition that the use of the lot method is improper where no attempts have been undertaken to subdivide the property is, once again, misplaced. The plaintiffs argue that the trial court, in the following excerpt from its decision, required the plaintiffs to take certain steps towards subdivision before it would validate the lot method. “Neither appraiser gave any consideration to the fact that . . . Langer owned this property since the early 1960’s and neither he nor his estate . . . made any effort whatever to undertake a subdivision of any part of the subject property. No application for a subdivision was ever filed nor any plan prepared, similar to Minicucci where [the] plaintiffs owned their property for 14 years and never attempted to undertake any subdivision of it.”
*409In order for a landowner to persuade the trier that his proposed method of valuation should be adopted, he must “provide the trier with sufficient evidence from which it could conclude that it is reasonably probable that the land to be taken would, but for the taking, be devoted to the proposed use by a prudent investor in the near future. . . . ‘The uses to be considered must be so reasonably probable as to have an effect on the present market value of the land. Purely imaginative or speculative value should not be considered.’ ” Tandet v. Urban Redevelopment Commission, 179 Conn. 293, 299-300, 426 A.2d 280 (1979).
While we agree with the plaintiffs that they need not take specific steps toward subdivision in order for the lot method to be considered, evidence of any attempts to prepare for a future subdivision tend to make such use more reasonably probable and less speculative. As the proponent of a hypothetical highest and best use is able to progress along the spectrum from raw land with few or no improvements to, ultimately, a completed subdivision, the weight to be assigned such evidence will be enhanced. 4 P. Nichols, Eminent Domain (3d Ed.) § 12B.14 [1] [b], pp. 12B-175-179. We conclude that the trial court did not mandate that any particular steps toward subdivision were required as a predicate to valuing the land on such a basis.
C
The plaintiffs next claim that because there was no dispute among the experts that the highest and best use of the property was as a subdivision and because there was no conflict about utilizing the lot method for appraising the land, the trial court was required to accept this method of determining fair market value and, therefore, it abused its discretion in rejecting the unanimous opinion of the four appraisers in their adoption of this method. We disagree.
*410“[T]rial courts must be afforded substantial discretion in choosing the most appropriate method of determining the value of a taken property. ... ‘In condemnation proceedings, the trial court is more than a trier of facts or an arbiter of . . . [the] opinions of [expert] witnesses; it is charged with the duty of making an independent determination of value and fair compensation in light of all the circumstances, the evidence, its general knowledge and its viewing of the premises.’ ” (Citations omitted.) French v. Clinton, 215 Conn. 197, 200-201, 575 A.2d 686 (1990). We conclude, therefore, that the trial court was not, as a matter of law, bound by the valuations or valuation methods used by the appraisers but could consider the comparable sales of land that were in evidence as well as the raw data utilized in the presentation of the lot method approach in independently determining fair market value. Second Stone Ridge Cooperative Corporation v. Bridgeport, 220 Conn. 335, 342, 597 A.2d 326 (1991) (an analysis of comparable sales is a valid method for determining the fair market value of property).
D
The plaintiffs next claim that, based on the facts of this case, the trial court improperly rejected the lot method of valuation because: (1) in French v. Clinton, supra, in which a conceptual marina was the highest and best use, this court approved of the lot method of valuation even though the condemned property was, with the exception of a small dwelling, undeveloped; and (2) the subject property was more amenable to subdivision than the condemned land in Minicucci v. Commissioner of Transportation, supra, and, therefore, the lot method should have been used. A review of the trial court’s findings with respect to the report of the plaintiffs’ appraisers discloses the following.
*411With respect to Glendinning’s report and his estimates of time delays for various town and state permits, the trial court concluded that the minimum delay would not be six months but rather at least one year. The trial court found that “[n]o traffic studies had yet been made. Drainage and detention basins had to be approved. The entire layout would require approvals by the defendant town and the State.” The trial court also determined that Glendinning’s estimate of less than four years to sell all the lots was too short and, like the other estimates, filled with a “substantial amount of speculation and prognostication of the future.” As to Glendinning’s attempt to compare an average lot in the hypothetical subdivision to comparable lots located elsewhere, and his separate evaluations of all forty-six lots, the trial court concluded that there had been “excessive partiality toward the subject residential zone” that did not enhance the credibility of his appraisal. The trial court further concluded that the estimated legal fees were too low and that it served “no useful purpose to discuss taxes, insurance, surety bonds, interest on construction financing based on further assumptions, interest on purchase mortgage, sales expenses, and profit and risk, roads, accessway[s], sewer and water mains, broker’s commissions, clearing and improving the land, taxes on unsold lots extending over 3% years minimum (based on Glendinning’s estimate), but likely to be considerably longer in the court’s judgment. The reason such discussion would not be useful is simply that it is based on speculation, conjecture and assumptions . . . [that have not been] quantified with reasonable exactitude and accuracy.” With respect to the report submitted by the plaintiffs’ other appraiser, Kennedy, the trial court determined that it was also not credible for many of the same reasons.
*412The trial court concluded that there was not “any reasonable probability the residential zone subject property would in fact be subdivided in the reasonably near future into any number of lots, whether 47 or less.” The trial court concluded, therefore, that the plaintiffs’ appraisers, by utilizing the lot method, had failed to value the land properly in its undeveloped state with some consideration allowed for its potential as a subdivision. The trial court thereafter valued the property at $9,425,000—a figure higher than the defendant’s appraisers’ value but lower than the compensation the plaintiffs had requested. We reiterate that a trial court “may seek aid in the testimony of experts, but must ultimately make its own ‘independent determination of fair compensation’; McDermott v. New Haven Redevelopment Agency, 184 Conn. 444, 447, 440 A.2d 168 (1981); on the basis of all the circumstances bearing upon value.” French v. Clinton, 215 Conn. 197, 202-203, 575 A.2d 686 (1990).
The plaintiffs argue that French represents the modern approach in eminent domain cases when undeveloped land is under consideration for valuation. In French, the condemned property consisted of 6.78 acres and contained a small cottage. “The trial court found that, due to the proximity of the property to Long Island Sound and the shortage of dockage and winter storage space available to the boating public, a small craft marina was the highest and best use of the premises. The [trial] court found in particular that a conceptual marina site plan proposed by the plaintiffs, which included 89 boat slips, year round storage for 100 boats and a winter storage shed for 50 boats, could be accommodated on the property.” French v. Clinton, supra, 199. The plaintiffs contend that since the still undeveloped marina was used in determining the fair market value of the property, the still undeveloped residential *413lots should have been similarly used to determine fair market value in this case. We disagree.
While French supports the proposition that the lot method of appraisal may be used in appropriate cases, in such cases, the proponent must offer credible evidence of the costs of the hypothetical subdivision and persuade the trier that, but for the taking, it would have occurred in the reasonably near future. Tandet v. Urban Redevelopment Commission, 179 Conn. 293, 299, 426 A.2d 280 (1979); see United States v. 47.3096 Acres, Etc., In Oxford Township, 583 F.2d 270, 271-72 (6th Cir. 1978). The plaintiffs failed to satisfy these threshold considerations for using the value of the completed lots in determining fair market value. Our holding in French v. Clinton, supra, does not mandate a different result.
The plaintiffs next argue that Minicucci does not reject, as a matter of law, the lot method, and that their property did not contain the types of visually determinable impediments to the development of a subdivision as found in that case. We agree. Minicucci v. Commissioner of Transportation, 211 Conn. 382, 384-86, 559 A.2d 216 (1989) (where undeveloped property contained a very steep slope, wetlands, rocky conditions and was heavily wooded and the landowners gave no hard evidence as to the costs of dealing with these impediments, the lot method was not an appropriate tool for appraising the property). The plaintiffs contend that because their land is better suited to development and they have offered evidence of the costs incident to preparing the land for a subdivision, the trial court should have adopted the lot method of valuation. We disagree. Despite the subject property’s greater adaptability to development, it was well within the trial court’s discretion, given the speculative nature of the evidence, to conclude that it was not “reasonably probable that the land would in fact be subdivided *414in the reasonably near future.” Minicucci v. Commissioner of Transportation, supra, 386. The trial court’s conclusion that, based upon the speculative evidence presented, the lot method was not an acceptable method of valuation was not clearly erroneous. Pandolphe’s Auto Parts, Inc. v. Manchester, 181 Conn. 217, 221, 435 A.2d 24 (1980).
The judgment is affirmed.
In this opinion the other justices concurred.
Although the plaintiffs seek a new trial on the value of the entire parcel, their appeal focuses solely on the value of the residentially zoned portion of the land.