The Petitioner, Citizens Association of Georgetown (“CAG”), noted an appeal from the District of Columbia Board of Zoning Adjustment (“BZA” or “Board”) order on remand approving the Georgetown University (“Georgetown” or “the University”) Campus Plan for Years 2000-2010 (“Campus Plan”). This matter is back before us after the BZA attempted to address the concerns raised in the original proceeding before this court. See President & Directors of Georgetown Coll. v. District of Columbia Bd. of Zoning Adjustment, 887 A.2d 58 (D.C. 2003) (“Georgetown I”). In that case, we vacated the BZA’s order and remanded the case for further proceedings after concluding that a condition imposed by the Board that froze the University’s full-time student enrollment was not supported by substantial evidence and that certain conditions, to which the University did not consent, were arbitrary and capricious. On remand, the BZA reconsidered the evidence presented and approved the Campus Plan with a revised cap and certain other conditions. Petitioner filed the instant petition for review seeking reversal of the latest BZA order approving the Campus Plan. Because there is substantial evidence in the record to support the BZA’s decision to increase the enrollment cap by averaging the number of full-time students enrolled during the fall and spring semesters, and that averaging for purposes of establishing an enrollment cap flows rationally from the BZA’s findings, we affirm the BZA order to the extent the parties have challenged that provision of the order. However, we again remand this matter to the BZA to explain why several uncontested provisions included in the Original Campus Plan were not included in the Revised Campus Plan issued in this case.

I.

FACTUAL AND PROCEDURAL BACKGROUND

On January 31, 2000, the University submitted an application for review and approval of its Campus Plan for Years 2000 to 2010 to the Board. On March 29, 2001, the Board conditionally approved the Campus Plan. Georgetown appealed the March 29, 2001 order. On December 4, 2003, we vacated the entire Original Campus order and remanded the case back to the Board for further proceedings. See Georgetown I, supra, 837 A.2d at 58.

In Georgetown I, we held that the Board’s order freezing the University’s enrollment cap at the levels set in 1990 was not supported by substantial evidence. Id. at 76. Specifically, we concluded that the BZA’s refusal to increase the enrollment cap from 5,627 to 6,016 lacked substantial evidence to support it, and ordered the BZA to “make reasonably detailed underlying evidentiary findings in which it specifically identifie[d] the need for continuing *588the 1990 cap and described] in non-conclu-sory terms the manner in which the retention of the cap would protect the residents of the adjoining communities.” Id. at 75. We also struck several other conditions imposed by the Board’s finding that those conditions were arbitrary and capricious. The matter was remanded to the Board for further proceedings consistent with our opinion. Id. at 83.

On June 22, 2004, in light of this court’s decision in Georgetown I, the Board held a public meeting and requested that the parties submit a list of issues to be addressed on remand. Thereafter, on October 15, 2004, the Board issued an order directing the parties to submit a proposed order either granting or denying the application in whole or in part, that included findings of fact, conclusions of law, and any proposed conditions that would mitigate any potential adverse impacts identified by the court’s decision in Georgetown I. On December 24, 2004, the parties submitted the requested documents. The CAG’s proposed order supported the Original Campus Plan. That plan included the following relevant conditions: maintaining the enrollment cap at 5,627 (Condition 2);1 mandatory reporting of data regarding off-campus student conduct (Condition 6); restricting the use of the Performing Arts Center, Harbin Field, and McDonough Arena (Condition 8); restricting the helipad for medically necessary purposes (Condition 9); requiring the University to include certain information in future Campus Plan applications (Condition 13); and requiring the University to submit periodic reports regarding its compliance with the Campus Plan (Condition 14). In addition to filing its proposed order, and findings of fact and conclusions of law, the CAG submitted data regarding off-campus student conduct occurring after the 2000 BZA proceedings, and a Zoning Commission order discussing the University’s use of averaging. The University’s proposed order included a Revised Condition 2, setting an enrollment cap of 6,016 full-time students. The proposed order made clear that the enrollment cap proposed by the University was arrived at by averaging the fall and spring traditional undergraduate numbers. The other conditions proposed by the University were identical to those in the proposed order by the CAG.

On February 1, 2005, the University filed a Motion to Strike the new evidence concerning the student data attached to the CAG’s proposed order. On April 5, 2005, the Board held a Public Meeting to discuss the proposed orders submitted by the parties on remand. At that time, the Board granted Georgetown’s Motion to Strike the new evidence attached to the CAG’s order. Thereafter, each member of the Board acknowledged on the record that he or she had reviewed the entire record of the Original Campus Plan proceeding. Subsequently on June 7, 2005, the Board issued its final order approving the Revised Campus Plan. That plan included the University’s Revised Condition 2, increasing the enrollment Cap to 6,016. In addition, the Revised Campus Plan eliminated certain conditions that were included in the Original Campus Plan and in the proposed orders submitted by both the CAG and the University. On June 9, 2005, the CAG filed a letter requesting the Zoning Commission (“Commission”) review, sua sponte, BZA’s Revised Campus Plan order. The Commission declined the CAG’s request.

*589ii.

LEGAL ANALYSIS

The Standard of Review

Generally, “[o]ur review of the Board’s factual determinations is deferential.” George Washington Univ. v. District of Columbia Bd. of Zoning Adjustment, 831 A.2d 921, 931 (D.C. 2003). We will uphold the Board’s findings “if they are based on substantial evidence as a whole.” See D.C.Code § 2-510(a) (2001); Georgetown Residents Alliance v. District of Columbia Bd. of Zoning Adjustment, 816 A.2d 41, 45 (D.C. 2003); Watergate West, Inc. v. District of Columbia Bd. of Zoning Adjustment, 815 A.2d 762, 765 (D.C. 2003). “Substantial evidence is relevant evidence which a reasonable trier of fact would find adequate to support a conclusion.” Giles v. District of Columbia Dep’t of Employment Servs., 758 A.2d 522, 524 (D.C. 2000). This court must determine (1) whether the agency made a finding of fact on each material contested issue of fact; (2) whether substantial evidence in the record supports each finding; and (3) whether the conclusions of law follow rationally from the findings. Foggy Bottom Ass’n v. District of Columbia Zoning Comm’n, 639 A.2d 578, 584-85 (D.C. 1994). Moreover, this court will defer to the Board’s findings and will not second-guess the Board’s decision unless it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Id.

The CAG contends (1) that the BZA erred when it allowed the University to average its fall and spring semester enrollment to determine compliance with the cap on undergraduate enrollment contained in the 2000-2010 Campus Plan, and that the BZA erred in failing to consider the new evidence of student data; and (2) that the BZA erred when it eliminated uncontested provisions from the Original Campus Plan Order.

III.

A Revised Condition 2 and Averaging

In Georgetown I, we concluded that the Board did not provide a justification for maintaining the University’s full-time enrollment cap of 5,627, which was the level set in the 1990 Campus Plan, instead of allowing the enrollment cap to increase to 6,016 upon the completion of the Southwest Quadrangle. 837 A.2d at 74-76. We noted that there was minimal “support in the record for the finding that the modest enrollment increase initially authorized but subsequently disapproved by the Board would have contributed to or exacerbated objectionable conditions in the adjoining neighborhoods.” Id. at 74. Thus, we held that the Board had to make detailed factual findings that would justify the cap of full-time student enrollment at the level set in the 1990 Campus Plan, and that the Board would have to describe how continuing the 1990 cap would protect the residents of the neighboring communities. Id. at 75.

In light of our remand order in Georgetown I, and after consideration of the original record as it existed in 2001,2 the Board concluded that the increase of the University’s enrollment cap to 6,016, which was calculated as an average over the fall and spring semesters of the academic year, would not become objectionable to *590neighboring property or have an adverse impact on the neighboring property. In reaching its determination, the Board relied on several factors, including evidence that the completion of the Southwest Quadrangle project would create housing for 84 percent of the traditional undergraduate population, and that the University had implemented new measures and enhanced existing programs to prevent and mitigate the impacts of any off-campus student misconduct in the neighborhoods surrounding the campus. The CAG now takes issue with the Board’s determination to increase the enrollment cap, but more vehemently challenges the Board’s decision to set a student enrollment cap based on an average of the University’s full-time undergraduate enrollment for the fall and spring semesters. According to the CAG, the increase in the enrollment cap along with the approval of the University’s averaging methodology, allows the University to “almost double” the number of undergraduate students enrolled in the fall semester. In essence, by permitting averaging, the University can increase its undergraduate student enrollment in the fall semester based on a lower enrollment in the spring semester, without exceeding its enrollment cap.

Our review is limited to determining whether there is substantial evidence in the record to support the Board’s findings of fact and whether the Board’s conclusions of law flow rationally from those facts. Foggy Bottom Ass’n, supra, 689 A.2d at 584-85. Thus, we will defer to the Board’s factual determinations. George Washington Univ., supra, 831 A.2d at 931. In this case, our review of the record reveals that the University presented substantial evidence regarding how it calculated its full-time student enrollment and how averaging the fall and spring enrollment figures is part of that calculation. During the June 13, 2000 Public Hearings on the Campus Plan, the University offered testimony that it calculated its total full-time undergraduate population by backing out non-traditional students, part-time students, and non-degree candidates, and then averaging the full-time undergraduate population, taking into account the difference in the number of traditional full-time students who are enrolled in the fall, but do not return for the spring semester.

During the October 2000 public hearing, the University also made part of the record a chart which indicated that it used an averaging methodology to calculate its student enrollment figures.3 Subsequently on October 6, 2000, in response to the Board’s request for clarification of its enrollment figures, the University submitted a letter and attached a chart discussing how it used averaging to calculate its full-time traditional undergraduate figure. The chart also set forth the rationale behind the University’s use of averaging to determine student enrollment:

The Fall enrollment number of traditional full-time undergraduate students is 5480. This compares with last year’s Fall enrollment number of 5744. Fall and Spring enrollment are averaged to reflect the academic year. Historically, Spring numbers range from 6-10% less than Fall numbers due to students studying abroad, mid-year graduation, withdrawals, etc. When the projected Spring numbers are averaged with the Fall number, the projection is 5603, which is less than the projection of 5677 previously included in Appendix E to the plan. (Emphasis added).

*591Given the substantial evidence in the record regarding the revised cap and the use of averaging as a methodology for calculating student enrollment caps, we see no basis to disturb the BZA’s decision in this regard.

Alternatively, the CAG contends that the Board’s decision to permit the University to average its fall and spring enrollment figures to determine an appropriate enrollment cap is arbitrary and capricious and thus, must be overturned. In essence, the CAG complains that the type of enrollment cap included in the Revised Campus Plan is not a true enrollment cap because it does not set a finite limit on the number of students that the University can enroll at any one time. While that is certainly another way of defining an enrollment cap, we are not persuaded that the enrollment cap imposed here does not flow rationally from the evidence presented or is otherwise unlawful. As we understand it, the main purpose of including an enrollment cap on the number of students a college or university can enroll as part of a campus plan is to limit the adverse impact the student population will have on the surrounding community. While we certainly appreciate why some limitation on enrollment has to be set, we see no reason why those limits have to be accomplished through the use of a hard cap based on a snapshot in time. In this case, the University explained clearly why it favored the setting of a cap based on the fluctuation it experiences in its seasonal enrollment figures. The Board also heard evidence from the CAG and the ANC about the impact that the University’s average number of traditional full-time students would have on the community and decided that the impacts were not likely to become objectionable or adversely affect the use of the neighboring property. In other words, the Board concluded that while there may be some difference in the enrollment numbers in the fall and spring, the differences between a hard enrollment cap and the blended enrollment cap advocated by the University would not have an adverse impact on the surrounding community. Because the Board’s decision in this regard seems neither arbitrary nor capricious, and “rationally flow[s] from findings of fact supported by substantial ¿vidence in the record as a whole,” we see no basis to disturb the Board’s ruling. See Watergate West, supra, at 765.

Finally, the CAG contends that despite its findings to the contrary, the Board failed to consider how the increase in the enrollment cap would affect the surrounding community. The Board’s findings of fact, however, indicate otherwise. With respect to the effects of averaging on the community, the Board found that

[The University’s] proposal to increase its enrollment cap on the number of traditional undergraduate students, calculated as an average over the Fall and Spring semesters of the academic year, is not likely to become objectionable to the neighboring property or to adversely affect the use of neighboring property. After completion of the new Southwest Quadrangle project, the University will have more than 5,000 beds on campus, a number sufficient to house 84 percent of the traditional undergraduate population. The University has implemented new measures and enhanced existing programs that will help to prevent and mitigate the impacts of any student misconduct off-campus in the neighborhood abutting the campus. (Emphasis added).

Thus, the record indicates that the Board considered how the increase in the enrollment cap would affect the neighboring property. Additionally, the Board’s findings also reflect that it gave “great weight” *592to the ANC’s requirement that the University maintain 85% of its undergraduate student population on-campus after completion of the Southwest Quadrangle project. See Foggy Bottom Ass’n v. Dist. of Columbia Bd. of Zoning Adjustment, 791 A.2d 64, 76-77 (D.C. 2002) (“ ‘Great weight’ implies explicit reference to each ANC issue and concern as such, as well as specific findings and conclusions with respect to each.”); 11 DMCR 3115.2 (2003) (“The Board shall give ‘great weight’ to the written report of the ANC [.]”). Specifically, the ANC “conditioned its approval of the proposed [Cjampus [P]lan” on the University’s recognition and BZA’s support that “measures be taken to strengthen the off-campus affairs program.”4 The Board’s consideration of the ANC’s requirements is also reflected in its order on remand noting that it gave “great weight” to the ANC. As a result, we are satisfied that the Board’s decision responded to and accorded “great weight” to ANC’s concerns. See Foggy Bottom Ass’n, supra, 791 A.2d at 77 (concluding that the Board gave “great weight” to the ANC by addressing the issues and concerns raised by the ANC with respect to the impact that a proposed hospital would have on the ANC’s constituents). Finally, because there is evidence in the record to support the Board’s findings and the Board’s conclusions flow rationally from those factual findings, we hold that the Board did not err in approving Revised Condition 2 of the Campus Plan.5

B. Removal of Uncontested Provisions

The CAG argues that the Board acted in an arbitrary and capricious manner when it omitted certain conditions that were included both in the Original Campus Plan and the proposed orders of the University and the CAG.6 We agree. In *593Georgetown I, we specifically discussed the conditions that we deemed to be arbitrary and capricious and directed the Board to “formulate appropriate conditions which are consistent with applicable legal requirements as set forth herein.”7 See Georgetown I, supra, 837 A.2d at 83. On remand, the Board removed the conditions that were uncontested in the original and remand proceedings. This court is troubled by the fact that the Board found the uncontested conditions necessary and supported by substantial evidence in the Original Campus Plan, but decided to eliminate them without explanation on remand. We find that the Board’s failure to give any findings or reasons for the omission of the uncontested conditions to be arbitrary and capricious.

Our decision in Levy v. District of Columbia Bd. of Zoning Adjustment, 570 A.2d 789, 752-53 (D.C. 1990), is applicable to the instant case. In Levy, the Board eliminated a condition on a motion for reconsideration in the same proceeding without explaining its findings. Id. at 752. We held that the unexplained elimination of the condition required reversal because “the modified order’s conclusion le[ft] the record in a state where the agency’s action not only d[id] not rationally flow from the findings, but [were] contrary to them.” Id. at 753. (internal citations and quotations omitted). We emphasized that our role was to “assure that the agency has given full and reasoned consideration to all material facts and issues.” Id. Similarly, we find here that the Board’s failure to revisit its earlier findings which were supported by substantial evidence, or to provide an explanation, for not including the uncontested conditions in the Campus Plan on remand, left the “record inadequate to reveal the basis for its decision” and was arbitrary and capricious. See id.

The University cites to Spring Valley-Wesley Heights Citizens Ass’n, et al., v. District of Columbia Zoning Comm’n, 856 A.2d 1174, 1181-82 (D.C. 2004), for the proposition that the Board has the discretion to eliminate conditions from a Campus Plan Order. But Spring Valley is a far different case from the one at bar. In Spring Valley, the court found that the Zoning Commission had the discretion to eliminate conditions from an earlier campus plan approved by the BZA. See id. Specifically, we concluded that the “Commission was not bound by what the BZA decided in a separate proceeding on an application for approval of a different campus plan.” Id. at 1181. In this case, however, the BZA reviewed its own decision and eliminated conditions that the Board previously found were supported by substantial evidence from the same campus plan, without explanation. Because the Board’s failure to provide an explanation for its decision to eliminate certain uncontested conditions included in its Original Campus Plan order makes it impossible to determine whether the decision is based on substantial evidence in the record, we must remand the record to the Board so that the basis for the exclusions can be ascertained.8

*594Accordingly, we conclude that there is substantial evidence in the record to support the BZA’s decision capping full-time student enrollment based on averaging the number of full-time students enrolled during the fall and spring semesters, and that averaging for purposes of establishing an enrollment cap flows rationally from the BZA’s findings. Thus, we affirm the BZA order to the extent the parties have challenged that provision of the order. However, we again remand the case back to the BZA for an explanation as to why several uncontested provisions included in the Original Campus Plan were not included in the Revised Campus Plan.

So ordered.