This is an appeal from the Superior Court’s (Cann, J.) dismissal of a petition for certiorari brought by the plaintiff pursuant to RSA 677:15, after the planning board of the town of Barn-stead (the board) denied the plaintiff’s subdivision proposal. We reverse and remand.
The board held a preliminary hearing on the plaintiff’s proposed subdivision on April 19, 1984. The proposal contemplates division of twenty-four acres into nine lots, each to be used for one triple-family *47townhouse unit with a separate water and septic system. The plaintiff owns the undeveloped land, formerly known as the “Old Fair property,” which is located in a residential zone on old Route 28.
Discussion at the hearing on the subdivision application included the impact the proposal would have on the town’s growth rate, as well as its impact on the school system, traffic, the construction and maintenance of suitable roadways, and the town’s future water supply. The board suggested that the plaintiff consider a five-lot subdivision with construction of single-family units phased in over five years, rather than the proposed twenty-seven units within one or two years. The plaintiff did not alter his proposal to reflect the board’s suggestion.
A second public hearing was held on May 17, 1984, at which the plaintiff resubmitted his proposal along with additional information in answer to questions raised at the first hearing. The plaintiff also agreed to phase construction, with twelve units to be completed during the first year, and the remaining fifteen during the second year. The board voted to continue the matter until the water supply and pollution control commission (WSPCC) approved the subdivision. Conditional approval was granted by the WSPCC on June 19, 1984, with the condition being that individual septic systems be approved at a later date after lot lines were approved and systems could be specifically located.
The board held its final hearing on June 21, 1984, and on September 5, 1984, the board met and made a final decision on the plaintiff’s proposal. The board unanimously voted to deny the subdivision proposal for the following reasons: (1) the impact the subdivision would have on the town’s growth rate; (2) its impact on the schools; and (3) “a concern for natural resources,” more specifically described as protection of water quality and the possible use of the site as a future water supply. The trial court upheld the board’s denial and dismissed the plaintiff’s petition for certiorari.
On appeal, the plaintiff contends that: (1) the trial court erred in ruling that a planning board may deny approval of a proposed subdivision on the basis of growth control rates set in a master plan, without the municipality’s first having enacted ordinances providing for growth management or the board’s having itself enacted a capital improvement program; (2) the evidence did not support the trial court’s finding that the three percent growth rate included in the master plan was based on scientific, statistical data; and (3) the trial court erred in upholding the board’s denial of the proposed subdivision in order to preserve a potential water supply, without providing just compensation to the plaintiff owner. We find that the *48trial court erred on the first and second issues and that the trial court lacked a basis for its third ruling.
We turn first to a discussion of Barnstead’s master plan. The board adopted a master plan for Barnstead and argues that it was properly applied as a “guideline” in the case at bar when the board found that the proposed subdivision would result in exceeding the town’s desired growth rate. The plaintiff counters that the planning board could not apply the master plan itself to limit development, but could only apply ordinances enacted pursuant to RSA 674:22 or RSA 674:23 to regulate or control the timing of development. The above-mentioned statutes enable a municipality to adopt an ordinance providing for controlled growth after its planning board has adopted a master plan and a capital improvement program designed to assess and balance community and regional development needs.
The plaintiff asserts that although the planning board has adopted a master plan, it has not adopted a capital improvement program, nor has any ordinance been enacted pursuant to RSA 674:22 or :23 which would provide guidance to a potential developer or applicant for a subdivision proposal. The plaintiff argues that the board cannot use the master plan in the absence of the town’s enacting ordinances in these areas. We agree.
According to statute, the master plan “shall generally be comprised of a report [and information]. . . designed to show as fully as is possible and practical the planning board’s recommendations for the desirable development of the territory legally and logically within its planning jurisdiction.” RSA 674:2 (emphasis added). If a town chooses to limit growth or to accomplish goals allowed by statute, then it must comply with the statutory scheme. The statutes as written provide that a town, through its elected legislative body, may adopt regulations or ordinances which provide for limited growth based on community and regional development needs, or may adopt interim regulations on a temporary basis. RSA 674:22 and :23. We have consistently upheld this interpretation of the statute and have further stated that a town must conform with the statutes which require adoption of a comprehensive plan to deal with long-term growth. Stoney-Brook Dev. Corp. v. Town of Fremont, 124 N.H. 583, 474 A.2d 561 (1984); Conway v. Town of Stratham, 120 N.H. 257, 414 A.2d 539 (1980).
In this case, if the board were allowed to apply the master plan directly, it would give greater legal status to the master plan than the legislature intended in RSA chapter 674. RSA 674:1 to :5 require a planning board to adopt a master plan but do not require *49the local legislative body to approve the plan. RSA 674:3. When a local legislative body enacts ordinances to regulate or control the timing of development, the public receives concrete guidelines. The enactments are subject to public and judicial scrutiny, and may be struck down if unlawful or unconstitutional. When a planning board purports to apply its limited growth recommendations on an ad hoc basis in place of limited growth legislation, it circumvents the requisite legislative process. We cannot countenance this practice.
In reaching its decision, the trial court compared New Hampshire law with the law of other jurisdictions relative to the legal effect of a municipality’s master plan and concluded that the board could deny a proposal solely under the aegis of the master plan. The court specifically cited the laws of Maryland, Florida, and Montana. The trial court’s reliance on law from these other jurisdictions is misplaced. In those jurisdictions, a master plan is the primary land use planning tool. Local legislative bodies, not simply planning boards, are required to adopt master plans, and thus, these plans acquire the force of law. New Hampshire law, on the other hand, states that the master plan should function as a guide in the land use planning process. Thus, the trial court erred in elevating the master plan’s status to that accorded master plans in the other jurisdictions considered.
Next, we turn to the plaintiff’s argument that the three percent annual growth rate adopted in the master plan did not have a “solid scientific, statistical basis.” Patenaude v. Town of Meredith, 118 N.H. 616, 621, 392 A.2d 582, 585 (1978). See also Stoney-Brook Dev. Corp. v. Town of Fremont, 124 N.H. at 588, 474 A.2d at 563. We have stated in the past that growth controls cannot be permanent or unreasonable. They must be carefully studied and continually reexamined in order to relax or eliminate them. See Beck v. Town of Raymond, 118 N.H. 793, 800, 394 A.2d 847, 852 (1978). Accordingly, planning boards must monitor growth and changes occurring within their communities. “Towns may not refuse to confront the future by building a moat around themselves and pulling up the drawbridge.” Id. at 801, 394 A.2d at 852.
In 1980, the office of State planning projected Barnstead’s population through the year 2000, according to a uniform projection process used for all State municipalities. The town’s actual population in 1980 was 2493, but the projection for that same year shows a population of less than 2300 persons. Further, the planning projection for 1990 indicates a population of 2831 persons and for 2000, a population of 3125 persons. According to our calculations, these figures translate approximately to a 2.3 percent average yearly growth between 1980 and 1990 and a one percent average yearly growth *50between 1990 and the year 2000. These projections were provided after an actual average yearly growth rate of ten percent between 1970 and 1980. It appears, then, that the figures projected by the office of State planning are unrealistic and do not reflect the actual growth experience in the town of Barnstead. If the board is using these projections to establish the ideal growth rate for the town, then it appears to us to be an unreasonable rate, as the plaintiff suggests. In addition, apparently the town also distributed questionnaires among Barnstead residents asking whether they regarded one, three or ten percent as the “ideal” yearly growth rate for the town. The three percent figure was most frequently selected. The master plan compared the figures and stated that “the State projection is close to what many townspeople feel to be a reasonable and moderate level of growth.”
The plaintiff offered expert testimony by a planning consultant who stated that a three percent growth rate was not a realistic approximation of actual growth. See Stoney-Brook, supra at 588, 474 A.2d at 563. The lower court found that the expert’s opinion was derived by studying geographical areas which were dissimilar to the town of Barnstead and, thus, had little relevance to Barnstead’s particular problems. Accordingly, the trial court gave little weight to the opinions of the expert planning consultant. We do not question this finding by the trial court. “The credibility and weight to be given to a witness’ testimony, is a question of fact for the trial court. If the findings can reasonably be made on all the evidence, they must stand.” Liberty Mut. Ins. Co. v. Custombilt, Inc., 128 N.H. 167, 512 A.2d 1098 (1986). We do question whether a finding by the trial court that the plaintiff’s expert’s testimony should be given little weight leads inexorably to the conclusion that the State projection is realistic. It is not clear to us that the town questionnaire combined with the State projection constitutes a solid, scientific, statistical basis for the recommended three percent growth rate. We hold that the evidence on the record was not sufficient for the trial court to reach the conclusion that the town’s decision was based on scientific, statistical data. See id.
We have held that towns, acting in good faith, “must develop plans to insure that municipal services, which normal growth will require, will be provided for in an orderly and rational manner.” Beck v. Town of Raymond, 118 N.H. at 801, 394 A.2d at 852. Hence, the validity of a town’s growth control ordinance rests on the relationship between the town’s restrictions on growth and a projection of what is deemed to be “normal growth” based on scientific and statistical evidence. Put simply, to date we have held that a *51growth control ordinance is valid only if it restricts projected normal growth no more than is necessary to allow for an orderly and good faith development of municipal services. The determination of what growth is normal rests, in part, on a consideration of the effects of zoning or other land use restrictions adopted for purposes of public benefit other than growth control as such.
An “orderly” and “good faith” growth control ordinance must in turn rest on what constitutes a reasonable rate of increase of municipal services for a particular town. Clearly, what is reasonable is a complex issue for which there can be no rigid test. See Burrows v. City of Keene, 121 N.H. 590, 598, 432 A.2d 15, 19-20 (1981). However, it seems clear that the test of reasonableness should include considerations of the cost of extending municipal services, the capacity of the town’s existing citizenry to adjust to the higher tax burden necessarily associated with an extension of municipal services, the probable use of the dwellings (e.g., recreational use as distinguished from primary homes), the availability and suitability of undeveloped land in neighboring towns and the overall growth of the region in which the town is located.
Hence, we hold that that “scientific and statistical basis” evidence of growth patterns describes but one kind of evidence bearing on the ultimate issue of the reasonableness of a growth control ordinance. Scientific and statistical evidence of growth projections cannot function as the sole guide as to what constitutes a reasonable growth limitation established by a particular town.
The final issue presented for our review is whether the trial court erred in upholding the denial of the subdivision proposal in order to preserve the area for a potential water supply. The plaintiff argues that if the town wants to hold his land undeveloped in order to save it for potential use as a water supply, he should be compensated for his inability to use his land. See N.H. Const. pt. I, art. 12; see also Funtown v. Town of Conway, 127 N.H. 312, 499 A.2d 1337 (1985). The town asserts that it has not denied the plaintiff the use of his land by denying his subdivision proposal because there are alternative uses available for the land, such as building single-family homes.
The trial court focused on the specific issue of septic systems in the plaintiff’s proposed development. Although the WSPCC gave its conditional approval to the plaintiff’s proposed project, the trial court was correct in stating that the planning board was not required to accept the WSPCC’s determination. Durant v. Town of Dunbarton, 121 N.H. 352, 357, 450 A.2d 140, 144 (1981). The trial *52court was correct in its reading of Durant and in its application of the case to the specific question of whether Barnstead must accept the WSPCC’s finding on septic systems. Obviously, under Durant the town is free to disregard the agency’s finding. See id.
The trial court erred, however, in not considering the town’s decision in this regard as a whole. We conclude that the board’s decision is arbitrary and unreasonable. There is no indication from the master plan that Barnstead has developed specific standards for aquifer protection. Nor has Barnstead designated what areas of the town are potential water supplies so that a purchaser will be aware of possible restrictions prior to buying such land. Since the town has not taken any action in this area, it had no basis for concluding that the plaintiff’s project would damage the water supply, that the building of a small number of single-family homes would safeguard the water supply, or that the WSPCC’s decision was incorrect. Since we order the board to reconsider its decision in this matter, we need not consider whether the town’s action amounted to an unconstitutional taking of property.
The conclusion we reach today does not result from a desire to dictate to local planning boards that particular decisions must be made, nor do we decide what the ultimate decision concerning the proposal in this case should be. Rather, we are concerned that planning boards adopt well-considered goals, apply their master plans consistently, and provide detailed explanation for a denial of a proposal. There must be an attempt by town and city planners to give greater guidance and more detailed rationales for the denial of a proposal. Inherent in this standard is a requirement that landowners and developers cooperate in good faith with the municipality.
The decision of the trial court is vacated and the case is remanded.
Vacated and remanded.
Thayer, J., did not sit; the others concurred.