"[T]he circumstances surrounding a checkpoint stop and search are far less intrusive than those attending a roving-patrol stop. Roving patrols often operate at night on seldom-traveled roads, and their approach may frighten motorists. At traffic checkpoints the motorist can see that other vehicles are being stopped, he can see visible signs of the officers' authority, and he is much less likely to be frightened or annoyed by the intrusion. 422 U. S., at 894-895.'" Martinez-Fuerte, 428 U. S., at 558. See also id, at 559. Here, checkpoints are selected pursuant to the guidelines, and uniformed police officers stop every approaching vehicle. The intrusion resulting from the brief stop at the sobriety checkpoint is for constitutional purposes indistinguishable from the checkpoint stops we upheld in Martinez-Fuerte. The Court of Appeals went on to consider as part of the balancing analysis the "effectiveness" of the proposed checkpoint program. Based on extensive testimony in the trial record, the court concluded that the checkpoint program failed the "effectiveness" part of the test, and that this failure materially discounted petitioners' strong interest in implementing the program. We think the Court of Appeals was wrong on this point as well. The actual language from Brown v. Texas, upon which the Michigan courts based their evaluation of "effectiveness," describes the balancing factor as "the degree to which the seizure advances the public interest." 443 U. S., at 51. This passage from Brown was not meant to transfer from politically accountable officials to the courts the decision as to which among reasonable alternative law enforcement techniques should be employed to deal with a serious public danger. Experts in police science might disagree over which of several methods of apprehending drunken drivers is preferrable as an ideal. But for purposes of Fourth Amendment analysis, the choice among such reasonable alternatives remains with the governmental officials who have a unique understanding of, and a responsibility for, limited public resources, including a finite number of police officers. Brown's rather general reference to "the degree to which the seizure advances the public interest" was derived, as the opinion makes clear, from the line of cases culminating in MartinezFuerte, supra. Neither Martinez-Fuerte nor Delaware v. Prouse, 440 U. S. 648 (1979), however, the two cases cited by the Court of Appeals as providing the basis for its "effectiveness" review, see 170 Mich. App., at 442, 429 N. W. 2d, at 183, supports the searching examination of "effectiveness" undertaken by the Michigan court. In Delaware v. Prouse, supra, we disapproved random stops made by Delaware Highway Patrol officers in an effort to apprehend unlicensed drivers and unsafe vehicles. We observed that no empirical evidence indicated that such stops would be an effective means of promoting roadway safety and said that "[i]t seems common sense that the percentage of all drivers on the road who are driving without a license is very small and that the number of licensed drivers who will be stopped in order to find one unlicensed operator will be large indeed." Id., at 659-660. We observed that the random stops involved the "kind of standardless and unconstrained discretion [which] is the evil the Court has discerned when in previous cases it has insisted that the discretion of the official in the field be circumscribed, at least to some extent." Id., at 661. We went on to state that our holding did not "cast doubt on the permissibility of roadside truck weigh-stations and inspection checkpoints, at which some vehicles may be subject to further detention for safety and regulatory inspection than are others." Id., at 663, n. 26. Unlike Prouse, this case involves neither a complete absence of empirical data nor a challenge to random highway stops. During the operation of the Saginaw County checkpoint, the detention of the 126 vehicles that entered the checkpoint resulted in the arrest of two drunken drivers. 444 BLACKMUN, J., concurring in judgment Stated as a percentage, approximately 1.6 percent of the drivers passing through the checkpoint were arrested for alcohol impairment. In addition, an expert witness testified at the trial that experience in other States demonstrated that, on the whole, sobriety checkpoints resulted in drunken driving arrests of around 1 percent of all motorists stopped. 170 Mich. App., at 441, 429 N. W. 2d, at 183. By way of comparison, the record from one of the consolidated cases in Martinez-Fuerte showed that in the associated checkpoint, illegal aliens were found in only 0.12 percent of the vehicles passing through the checkpoint. See 428 U. S., at 554. The ratio of illegal aliens detected to vehicles stopped (considering that on occasion two or more illegal aliens were found in a single vehicle) was approximately 0.5 percent. See ibid. We concluded that this "record . . . provides a rather complete picture of the effectiveness of the San Clemente checkpoint," ibid., and we sustained its constitutionality. We see no justification for a different conclusion here. In sum, the balance of the State's interest in preventing drunken driving, the extent to which this system can reasonably be said to advance that interest, and the degree of intrusion upon individual motorists who are briefly stopped, weighs in favor of the state program. We therefore hold that it is consistent with the Fourth Amendment. The judgment of the Michigan Court of Appeals is accordingly reversed, and the cause is remanded for further proceedings not inconsistent with this opinion. It is so ordered. JUSTICE BLACKMUN, concurring in the judgment. I concur only in the judgment. I fully agree with the Court's lamentations about the slaughter on our highways and about the dangers posed to almost everyone by the driver who is under the influence of alcohol or other drug. I add this comment only to remind the Court that it has been almost 20 years since, in Perez v. BRENNAN, J., dissenting 496 U. S. Campbell, 402 U. S. 637, 657 (1971), in writing for three others (no longer on the Court) and myself, I noted that the "slaughter on the highways of this Nation exceeds the death toll of all our wars," and that I detected "little genuine public concern about what takes place in our very midst and on our daily travel routes." See also Tate v. Short, 401 U. S. 395, 401 (1971) (concurring statement). And in the Appendix to my writing in Perez, 402 U. S., at 672, I set forth official figures to the effect that for the period from 1900 through 1969 motor-vehicle deaths in the United States exceeded the death toll of all our wars. I have little doubt that those figures, when supplemented for the two decades since 1969, would disclose an even more discouraging comparison. I am pleased, of course, that the Court is now stressing this tragic aspect of American life. See ante, at 451. JUSTICE BRENNAN, with whom JUSTICE MARSHALL joins, dissenting. Today, the Court rejects a Fourth Amendment challenge to a sobriety checkpoint policy in which police stop all cars and inspect all drivers for signs of intoxication without any individualized suspicion that a specific driver is intoxicated. The Court does so by balancing "the State's interest in preventing drunken driving, the extent to which this system can reasonably be said to advance that interest, and the degree of intrusion upon individual motorists who are briefly stopped." Ante, at 455. For the reasons stated by JUSTICE STEVENS in Parts I and II of his dissenting opinion, I agree that the Court misapplies that test by undervaluing the nature of the intrusion and exaggerating the law enforcement need to use the roadblocks to prevent drunken driving. See also United States v. Martinez-Fuerte, 428 U. S. 543, 567 (1976) (BRENNAN, J., dissenting). I write separately to express a few additional points. The majority opinion creates the impression that the Court generally engages in a balancing test in order to determine 444 BRENNAN, J., dissenting the constitutionality of all seizures, or at least those "dealing with police stops of motorists on public highways." Ante, at 450. This is not the case. In most cases, the police must possess probable cause for a seizure to be judged reasonable. See Dunaway v. New York, 442 U. S. 200, 209 (1979). Only when a seizure is “substantially less intrusive," id., at 210 (emphasis added), than a typical arrest is the general rule replaced by a balancing test. I agree with the Court that the initial stop of a car at a roadblock under the Michigan State Police sobriety checkpoint policy is sufficiently less intrusive than an arrest so that the reasonableness of the seizure may be judged, not by the presence of probable cause, but by balancing "the gravity of the public concerns served by the seizure, the degree to which the seizure advances the public interest, and the severity of the interference with individual liberty." Brown v. Texas, 443 U. S. 47, 51 (1979). But one searches the majority opinion in vain for any acknowledgment that the reason for employing the balancing test is that the seizure is minimally intrusive. Indeed, the opinion reads as if the minimal nature of the seizure ends rather than begins the inquiry into reasonableness. Once the Court establishes that the seizure is "slight," ante, at 451, it asserts without explanation that the balance "weighs in favor of the state program." Ante, at 455. The Court ignores the fact that in this class of minimally intrusive searches, we have generally required the government to prove that it had reasonable suspicion for a minimally intrusive seizure to be considered reasonable. See, e. g., Delaware v. Prouse, 440 U. S. 648, 661 (1979); United States v. Brignoni-Ponce, 422 U. S. 873, 882-883 (1975); Terry v. Ohio, 392 U. S. 1, 27 (1968). Some level of individualized suspicion is a core component of the protection the Fourth Amendment provides against arbitrary government action. See Prouse, supra, at 654-655; Martinez-Fuerte, supra, at 577 (BRENNAN, J., dissenting) ("Action based merely on |