*591OPINION OF THE COURT
Melvin L. Schweitzer, J.These claims arise from a July 30, 2002 accident in which William Dahl and his seven-year-old daughter, Alexandra Dahl, were struck by a motorcycle that had been proceeding northbound on the Wantagh State Parkway (WSP) while they were riding bicycles on a bicycle path adjacent to the parkway. The gravamen of the claims is that the accident was the result of negligence on the part of the State of New York in failing to install a guide rail to separate the parkway from the bicycle path.
William Dahl (hereinafter claimant) testified that he and his daughter had gone to Cedar Creek Park to ride on the bicycle path and rode south about three miles — to “just before the third bridge or third Wantagh bridge” (trial transcript at 20)— before turning around, heading back to the park. The bicycle path runs along the east side of the parkway, to claimant’s left as he and his daughter proceeded northbound. He stated that they passed two to three dozen people riding, walking or rollerblading on the path as they rode southbound.
Claimant estimated that there was about eight or nine feet between the edge of the parkway and the bicycle path. He noted that, other than at bridges and at the southern end of the path where it enters Jones Beach at a “very sharp curve into the parking lot” (id. at 21), there were no guide rails separating the bicycle path from the parkway. The path was paved blacktop and the area between the path and the parkway was grass and gravel. He described the area where the accident occurred— both the parkway and the bicycle path — as a “moderate curve” leading to a bridge.
As claimant and his daughter approached the bridge, they were struck by the errant motorcycle — which he stated he did not see or hear approaching — and knocked to the ground. Claimant was dragged several feet but did not lose consciousness. His daughter landed farther to the east in a wooded area.
Claimant testified that he was very familiar with the bicycle path because he had jogged on it several times a week for the previous IV2 years.
Claimants called three Department of Transportation (DOT) employees, and an expert witness, in support of their claim. David Glass, who has been with the DOT for 22 years, testified that he has been a supervisor in the DOT’s planning unit for *592six years and has been the bicycle and pedestrian planning coordinator for over 10 years.
Mr. Glass referred to a September 30, 1997 letter from a DOT engineer to State Senator Norman Levy, which addressed the proximity of the bicycle path to the vehicular traffic on the parkway, and the lack of a barrier:
“This is in response to your constituent’s concerns regarding bicycle safety on the Wantagh Parkway bicycle path. We conducted a study to determine if guide rail along the bicycle path is warranted. The study revealed that the bicycle path was designed in accordance with current American Association of State Highway and Transportation Officials (AASHTO) guidelines for bicycle facilities. Sufficient distance separates bicyclists from the adjacent roadway. In sections where this separation can not be maintained, such as on bridges, guide rail has been erected. Our review of the recent accident history for the subject section of the Wantagh Parkway does not reveal any accidents involving motor vehicles and bicyclists. Considering the parkway’s high traffic volumes and the intensity of the bike path use, the safety record speaks well for the design of both the parkway and the bicycle path.”
Harold Tarry a DOT design unit supervisor testified that the WSP bicycle path was constructed in 1975 and that the only places where guide rail was installed along the path were where it passed over bridges. He indicated that the location of the subject accident was about 500 feet south of the Seaman’s Island Creek Bridge, which was the northernmost of the three bridges between Jones Beach, to the south, and Cedar Creek Park, to the north. He estimated the distance between the parkway and the bicycle path at that location as being 10 to 12 feet.
Mr. Tarry stated that the parkway and the bicycle path in the subject area both proceeded on a curve with a radius of 1,800 feet, which he stated was not a sharp curve. Asked his opinion as to where barriers should be placed “with respect to bicycle paths in relation to a roadway” (id. at 88), he responded:
“[Wlhere I would look would be if you had — if there was a sharp curve where you had a history of accidents and you had a very high usage on it, I guess that would have to be more — well, more that I’ve seen here in this span .... I would have to have seen a large number of run off the road accidents at *593the times where pedestrians are out there and even then, I would have to consider what is the effect of putting the guiderail [sic] up there ... If I’m going to put the motorist, all of the people that drive along there at serious risk of injury, I have to know that there is also a serious risk of many other people being injured” (id. at 88-90).
Mr. Tarry testified that the average number of vehicles using the Wantagh State Parkway was 16,000 per day, higher in the summer and lower in the winter. Referring to a 2001 study of the WSP bicycle path, Mr. Tarry stated that the highest usage was on a Sunday in July where there were about 2,000 people who used the path.
Mr. Tarry testified that the items considered when installation of a guide rail was contemplated were whether there were fixed objects or other roadside hazards with which an errant vehicle could collide absent a rail. He noted that any time a guide rail is installed, it creates risks to vehicular traffic. He was asked if he thought that a barrier consisting of shrubbery could be appropriate adjacent to a bicycle path and he indicated that, in certain circumstances it would be, but the purpose of a shrub barrier would not be to contain errant vehicles but to delineate the roadway and the bicycle path. He stated the guidelines provided that if a roadway and a bicycle path were less than five feet apart, a shrub barrier could be considered to prevent cars from intentionally driving on the bicycle path and to prevent bicycles from riding on the road.
Mr. Tarry’s opinion was that the subject area of the parkway was safe, the curve was in compliance with existing standards, there was no need for any additional signs and there was no need for a guide rail. He was asked about a 1980 accident that had been referred to by claimant’s counsel, apparently involving a bicyclist being struck by an automobile, and he stated that, in addition to it being 20 years earlier than the subject accident, it occurred five miles north, in an area where he estimated that traffic volumes were five times greater. He testified no guide rails were installed at that location in response to that accident.
Mr. Tarry described a number of examples of bicycle lanes, including those along Veterans Highway in Hauppauge, on Route 25 going toward Orient Point and along Montauk Highway, as examples of bicycle lanes that are adjacent to vehicular traffic, without separation, that are common throughout the county, on roads “that have speed limits of thirty miles *594an hour to fifty-five miles per hour and they range in traffic volumes to roads that have as little as this road [referring to the Wantagh State Parkway in the area in question] and much more than this road” (id. at 151-152).
Chapter 10 of the New York State Department of Transportation Highway Design Manual provides that the primary warrants for the installation of guide rails are the degree of the embankment slope adjacent to the roadway and the presence of fixed objects or roadside hazards along the highway. Mr. Tarry identified roadside hazards as something like a utility pole, tree or bridge abutment, or a deep body of water. He testified that the only hazards of this kind along the WSP in the area in question were bridge abutments, and these are the only areas where guide rails were placed.
Mr. Tarry indicated that guide rails themselves are considered roadside hazards because they could increase the severity of accidents, and that DOT policy is first to see if the roadside can be made safe without the use of guide rails, and if that is not the case, they consider the placement of guide rails, starting with the most flexible systems and moving toward guide rails composed of stiffer materials.
Mr. Tarry was asked about an AASHTO1 publication relating to the development of bicycle facilities (exhibit 14) and he stated it was the primary reference that he used in designing bicycle lanes and paths. He testified that when a two-way bicycle path is located adjacent to a roadway, AASHTO suggests it be kept as far as possible from the roadway, but if it is not possible to keep the path five or more feet from the roadway, AASHTO suggests that some sort of barrier to “signify that they’re two separate facilities” be installed (trial transcript at 158). This is not a guide rail, a crashworthy barrier intended to redirect vehicles, but rather a visual barrier, such as shrubbery, intended to demarcate the road and the path and prevent intended movement of bicycles and pedestrians onto the road and motor vehicles onto the path.
According to Mr. Tarry, the Design Manual indicates that guide rails could be especially hazardous to motorcycles — that an automobile normally will just bounce off a guide rail but with a motorcycle, the impact is generally severe and often fatal.
*595When asked if sometimes there are circumstances when the installation of guide rails are considered for the protection of pedestrians, Mr. Tarry identified such a situation as follows:
“If you had an area that was on the outside of a sharp curve where you expected that people were going to be running off the road and you had a playground or a bus stop, something that continually had large groups of people there for a significant portion of the day ... or another example would be across from a T intersection . . . where somebody might try to make the light and not make the turn and ran off — again if there was something over there where a lot of pedestrians might congregate” (trial transcript at 172-173).
He stated that the evaluation in such a situation would take into account whether there was a history of accidents in the area and whether there was evidence of things such as scarred trees, skid marks or signs getting knocked over. He described the area in question herein as not being on a sharp curve, not containing a bus stop or playground or other location where large numbers of people congregate and not having any history of accidents other than the one involving the claimants herein.
Frank Pearson, the director of the Traffic, Engineering and Safety Office of the BOTs Long Island Region, testified that the average daily traffic on this section of the parkway was about 11,000 cars, in both directions, which he stated was the lowest volume section of the parkway. He estimated that, north of Merrick Road, volume on the WSP increased to about 20,000 cars daily, and north of the Southern State Parkway, volume on the WSP was about 70,000 ears daily. He also estimated that daily traffic volume on the Veterans Memorial Highway, which has bicycle lanes adjacent to the highway without separation from vehicular traffic, was about 35,000 to 40,000 and that volume on the Cross Island Parkway in Queens, which was identified as containing a bicycle path that is separated from traffic by a barrier, was well over 100,000 daily.
Mr. Pearson stated that other than the accident involving the claimants and the accident which occurred in 1980 three or four miles away, he was not aware of any accident involving a collision between a motor vehicle and a pedestrian or bicyclist. Over the five-year period, there were no reported accidents of any kind at the specific location of the subject accident (exhibit O).
Mr. Pearson indicated that except where the existence of a fixed object dictated installation of a guide rail, the presence of *596the rail generally led to an increase in the number of accidents in an area: “if there is no object being shielded, they’re going to strike the [guide rail]. It would certainly be much more severe than running off the road and recovering” (trial transcript at 283).
Mr. Pearson conceded, with respect to the absence of prior accidents at the subject location, that if a vehicle left the roadway and crossed over the bicycle path, but did not strike anything, no accident report would be generated. He acknowledged that the WSP bicycle path is within the highway’s clear zone — the area adjacent to the roadway within which errant vehicles may recover from a loss of control — but stated that there are no specific standards that govern bicycle paths within clear zones. He advised that when considering whether to install a guide rail on a highway, the consideration is “what’s worse? Is it to hit that object off the side of the road or is it to hit the [guide rail?]” (id. at 303).
Lance Robson, a licensed civil engineer, was retained by claimants in October 2002, and was provided with the police accident report, photographs of the site and transcripts of the depositions of Mr. Tarry and Mr. Pearson. He went to the site on November 2, 2002, and took photographs (exhibit 1) and measurements. He measured the distance between the edge of the roadway and the edge of the bicycle path in the subject area as nine feet six inches, and the width of the path as ten feet nine inches. The edge of the right travel lane is separated from the path by grass, and there is no curb between the travel lane and the adjoining grassy area.
Mr. Robson referred to a 1995 revision of the Highway Design Manual (Engineering Instruction 95-013, revising ch 10 of the Manual [Roadside Design, Guide Rail and Appurtenances] [exhibit 9]) that discusses the “Developed Area and Large Volume Exceptions” (id.). The revision states that it applies to “[projects initiated after April 3, 1995 or with a design approval date after January 1, 1996” (id.). Asked how this revision related to his analysis of the WSP and the bicycle path, Mr. Robson noted the second paragraph, which states: “Finally, while in rural areas attention may be focused primarily on protecting the motorist from the roadside, in populated areas, consideration must also be given to protecting pedestrians and bicyclists from motorists” (id.). He stated that this was the first time that this subject was addressed in the manual. The revision also states: “Barriers should be considered where areas of assembly, *597particularly playgrounds, schools and parks, are across ‘T’ intersections, outside of sharp curves, and at locations with a history of run-off-road accidents” (id.).
After referring to this 1995 revision of the Highway Design Manual, and the 1989 AASHTO Roadside Design Guide quoted above (exhibit 10), Mr. Robson was asked whether the separation between the parkway and the bicycle path was adequate. He replied that it was not adequate because the “clear zone” in this location should have been more than 28 feet and thus the bicycle path was “squarely within the clear zone” (trial transcript at 367). His opinion was that the presence of the bicycle path in this location, 9V2 feet from the parkway, without a barrier, constituted a dangerous condition, “because of the numbers of people using it” (id. at 368-369), referring to the 2,076 daily users shown on exhibit “6A” on July 29, 2001. He acknowledged that such was the highest daily usage shown on the count, and stated that “during the winter, this would not be a dangerous location” (id. at 370), but contended that “if you have that danger for a month’s period or a two or three [months’] period, you have to deal with the danger” (id. at 371).2
Section 10.2.1 of the Highway Design Manual (exhibit 13) discusses the concept of the “clear zone,” which Mr. Robson testified became part of highway engineering practice in the late 1960s, based on the recognition that vehicles will run off the road from time to time. The manual states that the purpose of this section is to “provide guidance for the design of roadside features on new construction or reconstruction projects” (id.). Making reference to this section, Mr. Robson stated that the appropriate clear zone for the area of the WSP in question herein is 23 feet, thus the bicycle path is within the clear zone. Mr. Robson also noted that the Highway Design Manual considers a shoulder to be part of the roadway (see exhibit 15). Asked what was the significance of that concept to the issues in this case, he referred to the AASHTO Guide for the Development of Bicycle Facilities that Mr. Tarry had been asked about (exhibit 14), which states that where the separation between a bicycle path and an adjacent highway is less than five feet, a “suitable physical barrier may be considered” in order to “prevent bicyclists *598from making unwanted movements between the path and the highway shoulder and to reinforce the concept that the bicycle path is an independent facility” (id.). He asserted that shoulders are usually 8 to 10 feet wide and, therefore, that the appropriate measurement in this case, for purposes of considering the AASHTO guideline, should have started 8 to 10 feet from the edge of the roadway and that the “edge of this bicycle path is either in the shoulder or only two feet from the shoulder” (trial transcript at 397).
Mr. Robson stated he was not aware whether the parkway was in compliance with relevant standards when it was first constructed in 1928 or when it was reconstructed over the years, but he had no reason to think that it was not in compliance. He agreed the subject location did not meet any of the applicable warrants for the installation of a guide rail, a guide rail itself is a roadside hazard, and whether or not to install a guide rail should be the product of careful study because it involves a safety trade-off. Referring to the AASHTO Guide for the Development of Bicycle Facilities (exhibit 14), he conceded that the problem addressed therein arising from a bicycle path close to a roadway was the perception problem that arises in such situations, and the barriers referred to therein are the delineation barriers — fencing or shrubbery — that are recommended when a bicycle path is less than five feet from a roadway. He acknowledged “there’s nothing in this guide” referring to barriers intended to keep errant vehicles from a bicycle path (trial transcript at 431).
Defendant recalled Mr. Tarry as its sole witness. He was asked if there was anything in the Highway Design Manual that prohibits pedestrians, bicycles, bicycle paths or bicycle routes from being within a highway’s clear zone and he responded by stating that while it would be “great” if you could always have 30 or more feet of separation, it is “often” the case that areas intended for pedestrians and bicycles are within the clear zone (id. at 451).
Mr. Tarry also noted there is no requirement that highways have shoulders, and on highways without shoulders, ascertaining the distance from the roadway involves measuring from the edge of the roadway, or perhaps from a wide stripe or curb if such exists, and not adding some distance for a hypothetical shoulder that is not present. He testified there is “no direct guidance in the literature on [guide rail] between a path and a roadway” (id. at 471) — a statement that is uncontradicted by *599any of the evidence produced at this trial; and that when designing a new bicycle path, “what we end up doing is looking at each case on the merits and I would look at AASHTO, . . . look at the design manual, look at other publications . . . the traffic volumes, the path, proposed path usage volumes . . . and then I’d come up with a recommendation” (id. at 471-472).
The State has the duty to design, construct and maintain its roads so that they are reasonably safe (Friedman v State of New York, 67 NY2d 271 [1986]) and this duty applies as well to the design, construction and maintenance of guide rails (Lattanzi v State of New York, 53 NY2d 1045 [1981]). Although the State is not the insurer of the safety of those who use its roadways (Atkinson v County of Oneida, 77 AD2d 257 [1980]; Boyce Motor Lines v State of New York, 280 App Div 693 [1952], affd 306 NY 801 [1954]), the State’s duty of reasonable care does include the duty to install a guide rail when it is on notice of a dangerous condition that could have been addressed by the installation of a guide rail and where the failure to have done so was a proximate cause of injury (McDonald v State of New York, 307 AD2d 687 [3d Dept 2003]).
Cases addressing whether the State, or a municipality, is liable for alleged negligence in the design or construction of highways, including those addressing whether the State breached its duty to alleviate a known hazardous highway condition, generally revolve around the concept of “qualified immunity” (see Friedman v State of New York, supra; Alexander v Eldred, 63 NY2d 460 [1984]; Weiss v Fote, 7 NY2d 579 [1960]). Pursuant to this doctrine, a decision made by state highway planners may be successfully challenged in a negligence action only where it is shown that the decision was made without adequate study or lacked a reasonable basis (id.), where there was an unreasonable delay in implementing the decision (Carroll v State of New York, 157 AD2d 697 [1990]), or where the State failed in its duty to review and monitor the results and consequences of its decision (Friedman, supra, 67 NY2d 271, 284 [1986]; Atkinson v County of Oneida, supra, 77 AD2d 257, 261 [1980]).
Claimants argued strenuously, both at trial and in their post-trial submission, that the State was not entitled to rely on the qualified immunity defense because it failed to produce, either at trial or during disclosure, the studies that led to its decision not to install a guide rail at the subject location. Although various items or correspondence referred to warrant studies having *600been performed, the records of such studies could not be located and ¿1 the court was presented with were statements, both in the letters and at trial, that the warrants for the installation of guide rails had not been met. Claimant is correct in contending that such is insufficient to support a finding of qualified immunity (see e.g. Cordero v City of New York, 112 AD2d 914, 915 [2d Dept 1985] [“The city cannot rely on the doctrine of Weiss v. Fote (7 NY2d 579) merely by submitting conclusory declarations that surveys were made without at least producing written copies or testimony by those who made the surveys”]). However, the doctrine of qualified immunity is inapplicable herein for another, more basic, reason: claimants failed to prove a prima facie case of negligence, rendering defendant’s failure to establish a basis for qualified immunity irrelevant.
Although claimant’s expert opined that defendant “should have installed [a] barrier at the crash site to protect bicyclists and pedestrians from the foreseeable consequences of a vehicle leaving the outside of the curve” (trial transcript at 400), the record at this trial did not support this conclusion. The record also did not support claimants’ references, in their posttrial submission, to the State’s “violation of the AASHTO standards” (claimants’ posttrial brief at 58), since no such violation was demonstrated.
The 1991 AASHTO Guide for the Development of Bicycle Facilities (exhibit 14) referred to the visual problem that sometimes results when a bicycle path is too close to a roadway and suggests that when the separation is less than five feet, “a suitable physical divider may be considered . . . both to prevent bicyclists from making unwanted movements between the path and the highway shoulder and to reinforce the concept that the bicycle path is an independent facility” (id. at 24). Mr. Robson conceded that what is contemplated by this suggestion is not a crashworthy barrier intended to redirect errant vehicles, but rather something like a row of shrubs or a fence, intended solely for delineation purposes. His testimony that the State somehow misinterpreted this guideline because “there isn’t any shoulder along this roadway . . . [s]o the five feet should be taken from the edge of the shoulder” (trial transcript at 398) was not only self-contradictory, but also irrelevant to the issue herein, which is whether it was negligence not to have a crashworthy barrier (i.e., a guide rail) at this location. Indeed, the very existence of a guideline intended to address a visual problem that sometimes occurs when a bicycle path is within five feet of a roadway as*601sumes there is no guide rail in place, because a guide rail would obviate any need for a delineation barrier.
The 1989 AASHTO Roadside Design Guide (§ 5.2.3 at 5-4 [exhibit 10]), addressed the situation where protection of “innocent bystanders” might be necessary under circumstances where the conventional criteria did not warrant a guide rail. After noting “there are no objective criteria to draw on for pedestrian and cyclist barrier warrants,” AASHTO merely states that “when sidewalks or bicycle paths are adjacent to the traveled way of high speed facilities, some provision might be made for the safety of pedestrians and cyclists” (id. at 5-5). A statement that in certain situations, some provision might be made for something does not, in the absence of something more, create a standard that is violated, as claimants contend. On the same page of the excerpt relied on by claimants, AASHTO notes, with respect to traditional barrier warrants, that “[m]arginal situations, with respect to placement or omission of a barrier, will usually be decided by accident experience, either at the site or at a comparable site” {id.).
Similarly, the 1995 revision to chapter 10 of the Highway Design Manual (exhibit 9), which applies to projects initiated after April 1995, and which Mr. Robson acknowledged as reflecting for the first time the subject of protecting pedestrians in the manual, stated that “[b]arriers should be considered where areas of assembly, particularly playgrounds, schools and parks, are across ‘T’ intersections, outside of sharp curves, and at locations with a history of run-off-road accidents” {id.). Leaving aside that this instruction applies to new construction, not to existing facilities (see e.g. Stuart-Bullock v State of New York, 38 AD2d 626 [3d Dept 1971], affd 33 NY2d 418 [1974]), and that it, like the AASHTO guidelines, does not contain any mandatory direction to install a barrier but merely to consider it in certain situations, the parkway and bicycle path at issue herein do not fall into any of the categories referenced in the revision. This was not an area of assembly across from a “T” intersection or outside of a sharp curve3 and, most significant, there was no history of accidents involving vehicles running off the parkway at any portion along the bicycle path, a distance of more than four miles.
As has been noted in many decisions involving allegations of negligent highway design, construction or maintenance, the *602lack of any proof of prior similar accidents “is some indication at least that the highway was reasonably safe for those who exercised reasonable care” (Stapleton v State of New York, 285 App Div 984, 985 [3d Dept 1955]; see also e.g. Light v State of New York, 250 AD2d 988 [3d Dept 1998]; Galvin v State of New York, 245 AD2d 418 [2d Dept 1997]; Hough v State of New York, 203 AD2d 736 [3d Dept 1994]; Whiter v State of New York, 148 AD2d 825 [3d Dept 1989]; Boyce Motor Lines v State of New York, 280 App Div 693 [3d Dept 1952]).
Claimants’ contention that the absence of reports of prior accidents did not necessarily indicate the area was safe because this does not account for the possibility that vehicles may have gone off the road and made a successful recovery in the clear zone (perhaps passing over the bicycle path at a time when there were no bicyclists or pedestrians there and thus not resulting in an accident), actually lent support to defendant’s decision that no guide rail was required. Had a guide rail been in place, all such incidents, if there were any, necessarily would have resulted in collisions with the barrier, collisions that did not happen only because a barrier was not present. Claimants’ expert recognized that installation of guide rails necessarily involved a balancing of relative hazards, a balancing that, on this record, was appropriate.
Comparison of the decisions in McDonald v State of New York (307 AD2d 687 [3d Dept 2003]) and Chunhye Kang-Kim v City of New York (29 AD3d 57 [1st Dept 2006]) is instructive. In McDonald, a “T” intersection case, the Third Department held there was ample support for the Court of Claims’ finding that the State
“failed to maintain this stretch of guide rail in a reasonably safe condition for at least two months prior to the accident and that, based on the numerous accidents that occurred in this area through the years . . . defendant had actual or constructive notice of this dangerous condition and failed to take reasonable measure to remedy it” (307 AD2d 687, 688 [2003]).
In Chunhye Kang-Kim, where plaintiff alleged the City was negligent for failing to erect a barrier to protect pedestrians at an allegedly dangerous intersection, the First Department, quoting Hough v State of New York (supra at 738), noted that it was “incumbent” on the plaintiff to establish notice of a dangerous condition by producing evidence of prior similar accidents, and stated:
“This case differs from those cases relied upon by *603plaintiff in which a failure by the State or municipality to install guard rails or barriers was found to constitute negligence. In all of those cases the need for such safety devices was necessitated by a known dangerous condition or a prior history of accidents at the site” (29 AD3d 57, 60, 61 [2006]).
This language is equally applicable herein, as is the Court’s observation that the expert’s conclusions were not supported by reference to any standard requiring a barrier at the scene of the accident.
Claimants failed to sustain their burden of proving that any negligence on the part of the State of New York contributed to the causation of this accident. Accordingly, the Clerk of the Court is directed to enter judgment dismissing these claims.
. American Association of State Highway and Transportation Officials.
. Other than the count of 2,076 on Sunday, July 29, 2001, Mr. Robson did not address the actual counts for each date in the week from Wednesday, July 25, 2001 through Wednesday, August, 1, 2001: 587, 231,1171,1071, 2076, 637, 1296 and 647.
. Notwithstanding claimants’ irrelevant and inaccurate attempt to characterize the bicycle path as a “park.”