Skip Navigation LinksSkip Navigation Links
Centers for Disease Control and Prevention
Safer Healthier People
Blue White
Blue White
bottom curve
CDC Home Search Health Topics A-Z spacer spacer
Blue curve MMWR spacer

Laboratory Performance Evaluation of N95 Filtering Facepiece Respirators, 1996

In 1995, CDC's National Institute for Occupational Safety and Health (NIOSH) introduced a new classification scheme for particulate air-purifying respirators (1). Most health-care workers use type N95 half-mask filtering facepiece respirators (i.e., N95 respirators) to prevent occupational transmission of tuberculosis.* As a result, NIOSH received inquiries about how well N95 respirators fit, whether they need to be fit tested, and whether they can be quantitatively fit tested.** In response to these inquiries, NIOSH evaluated the performance of 21 N95 respirator models on a 25-person panel. This report summarizes the results of this evaluation, which indicate that fit testing is needed to ensure at least the expected level of protection (i.e., the concentration of airborne contaminants inside the respirator is less than or equal to 10% of ambient levels).

The panel comprised 15 women and 10 men (all experienced in wearing respirators and fit testing); the distribution of face lengths and face widths approximated that of the general population (2). The 21 respirator models were the only respirators commercially available in July 1996, when the evaluation began.

Each respirator model was assessed by 1) the 25-person panel without fit testing and 2) removing from the panel those persons for whom a model failed a surrogate fit test. For each model, total penetration (i.e., direct penetration through the filter and leakage around the face seal*** combined) was measured with each person on the panel using the TSI 8020 Portacount PlusTM****, a fit-test instrument that uses ambient air particles as the challenge agent (3). In a previous study, fit factors (the reciprocal of face-seal leakage) measured by this instrument correlated with actual exposure (4).

For each test, the person donned the respirator and performed a user seal check (i.e., pressure-tightness test, fit check, or negative/positive pressure check) according to the manufacturer's instructions; when respirator models were available in multiple sizes, the size with the best subjective fit was used. Each person then performed a six-exercise***** test during which total respirator penetration was measured. These exercises, each lasting approximately 80 seconds, simulate facial movements during normal use and typically are included in fit testing protocols. After removing the respirator, three identical repeat tests were performed. Total penetration was measured during each test; thus, four total penetration measurements were obtained with each respirator for each of the 25 persons.

For each respirator model, the resulting 100 total penetrations were used to calculate the 95th percentile of the total penetration, using the geometric mean (GM) and the geometric standard deviation (GSD) of these measures, as GM X GSD1.645 (5). These results summarize the performance of these 21 models without fit testing. Values for the 95th percentile ranged from 6% to 88% total penetration. Five respirator models had 95th percentiles of less than or equal to 10% total penetration (Table_1). The computed figure indicates that 95% of wearers of that model can expect a total respirator penetration less than this value and is used to indicate overall respirator performance (6).

For each person-respirator model combination, the first total penetration measurement then was used as a surrogate fit test to estimate N95 respirator performance when fit testing is conducted before use. Because fit tests are intended to assess only face-seal leakage, the measured total penetration was adjusted by subtracting the filter penetration,****** measured separately on each respirator by using the PortacountTM with a specially designed fixture. Each respirator having face-seal leakage greater than 1%******* during the first trial was considered to have failed the fit test for that person, and data for that person's trials were then removed from the data set for that respirator model (2). For respirators passing this criterion, total penetrations measured for trials 2, 3, and 4 were used to calculate the 95th percentile of the total penetration. These values summarize the performance of the respirators after a fit test was used to screen out respirators that have face-seal leakage greater than 1% (Table_1). The total penetrations ranged from 1% to 16%. For three models, none of the respirators passed the fit test (i.e., none had a first-donning face fit leakage less than or equal to 1%); therefore, the 95th percentile could not be computed. By applying the surrogate fit test, 17 of the 21 models had total penetration values less than or equal to 10%, a substantial increase in protection. Many models had a high fit test failure rate; 17 had acceptable fit tests for fewer than half of the panel members (Table_1).

Reported by: Laboratory Investigations Br, Div of Respiratory Disease Studies, National Institute for Occupational Safety and Health, CDC.

Editorial Note

Editorial Note: The findings in this report indicate that fit testing N95 respirators is essential in programs employing these respirators and can eliminate poorly fitting respirators, ensuring at least the expected level of protection. Without surrogate fit testing, average exposure for the 25-person panel was reduced to 33% of the ambient level, which is much less protection than expected of this class of respirators (i.e., exposure reduced to less than or equal to 10% of ambient levels). However, when fit tested first, the panel received substantially greater protection than normally expected (the average exposure was reduced to 4% of the ambient level). Without fit testing, persons unknowingly may have poor face seals, resulting in excessive leakage and exposure. For example, the respirators in this study had high fit test failure rates, with 20%-100% of panel members unable to achieve a satisfactory fit with a given respirator model.

The PortacountTM fit test instrument measures the large number (several thousand per cubic centimeter) of small particles present in normal room atmospheres. The instrument counts the number of such particles that penetrate the respirator -- either through face-seal leakage or directly through the filter. Previously, this instrument was recommended for use only with high-efficiency respirators that had negligible filter penetration because any particles detected inside the facepiece could be attributed to face-seal leakage. This study tested N95 respirators using the same procedure. However, because N95 filters are not 100% efficient in removing ambient air particles, two additional were steps needed:

  1. separate measuring of filter penetration and 2) subtracting this filter penetration (2). The technique for quantitatively fit testing N95 respirators in this report is appropriate only for research purposes. The manufacturer has recently developed an accessory to test N95 respirators with the Portacount PlusTM; the accessory removes the aerosols in the range that is most penetrating to the respirator filter, so filter penetration is not a concern. The approach used in this study suggests the possibility of commercial adaptation of similar fit test systems, resulting in a second, inexpensive means of quantitative fit testing N95 respirators. The availability of such a fit test system could simplify fit testing and would provide an option to persons responsible for overseeing respirator programs, especially those who already have the basic hardware for quantitative fit testing.

Although some models had (95th percentile) total penetrations less than or equal to 10% even without fit testing, these models should be fit tested. The findings in this report indicate that the models evaluated do not provide the expected level of protection for every user. Therefore, even for these models, performing a fit test has value in identifying those wearers having poor fit.

The findings in this study are subject to at least two limitations. First, specific models used do not necessarily represent the models now available; many are no longer marketed in the version tested, and continued product modifications by the respirator manufacturer may affect the fitting characteristics of specific models. Second, some models tested have been replaced with newer versions, and additional models are now available.

The fit test pass/fail level of 1% used in this report typically is recommended by respirator authorities (7). This criterion, however, is based on professional judgement. NIOSH will further analyze these data to determine the effect of adjusted pass/fail levels. Such analysis may provide insight into the appropriateness of that pass/fail level.


  1. Office of the Federal Register. Code of federal regulations: part 84 -- approval of respiratory protective devices. Washington, DC: Office of the Federal Register, National Archives and Records Administration, 1996. (42 CFR part 84).

  2. Coffey C. Testing of half-mask filtering facepiece N95 respirators with the TSI PortacountTM. Morgantown, West Virginia: US Department of Health and Human Services, CDC, National Institute for Occupational Safety and Health, 1998.

  3. TSI, Inc. Portacount PlusTM manual. St. Paul, Minnesota: TSI, Inc, 1994.

  4. Coffey C, Campbell DL, Myers WR, Zhuang Z. Comparison of six respirator fit-test methods with an actual measurement of exposure: part II -- method comparison testing. Am Ind Hyg Assoc J (in press).

  5. Lenhart SW, Campbell DL. Assigned protection factors for two respirator types based upon workplace performance testing. Ann Occup Hyg 1984;28:173-82.

  6. National Institute for Occupational Safety and Health. NIOSH respirator decision logic. Cincinnati, Ohio: US Department of Health and Human Services, Public Health Service, CDC, 1987:13-8; DHHS publication no. (NIOSH)87-108.

  7. American National Standards Institute. American national standard for respiratory protection, ANSI Z88.2-1992. New York, New York: American National Standards Institute, 1992.

* There are nine classes of filters (three classes of filter efficiency {95%, 99%, and 99.97%} each with three categories of resistance to filter efficiency degradation {N, R, and P}). N-category filters are the least resistant to degradation by oil aerosols. An N95 filter is an N-category filter that is at least 95% efficient.

** Fit testing is a procedure used to evaluate how well a given respirator fits a given person by assessing leakage around the face seal; fit testing can either be qualitative (i.e., relying on a subjective response of the wearer) or quantitative (i.e., using a measurement of actual leakage).

*** PT=Pfp+Pfsl, where PT is the total penetration, Pfp is filter penetration, and Pfsl is face-seal leakage.

**** Use of trade names and commercial sources is for identification only and does not imply endorsement by the U.S. Department of Health and Human Services or CDC.

***** Normal breathing, deep breathing, moving head side to side, moving the head up and down, reading a prepared text aloud, and normal breathing.

****** Pfsl=PT-Pfp.

******* The 1% criterion is the standard value used by the Occupational Safety and Health Administration and the American National Standards Institute to assess face seal leakage and is intended to provide a 10-fold safety factor between laboratory-based assessments of leakage and leakage during actual working conditions (i.e., less than 1% leakage in the lab should assure less than 10% leakage in the field).

Note: To print large tables and graphs users may have to change their printer settings to landscape and use a small font size.

TABLE 1. Performance testing data for 21 N95 filtering facepiece respirators for 25
persons, 1996
                                                     passing surrogate
                                                         fit test
Respirator         Total penetration*                           Total penetration*
model              (95th percentile)+       No. passing         (95th percentile)+
  1                        6%                    14                     3%
  2                        7%                     8                     2%
  3                       18%                    16                     1%
  4                       88%                     4                     1%
  5                       31%                     0                     NA&
  6                       11%                    15                     4%
  7                       10%                     5                     2%
  8                        6%                     9                     2%
  9                       18%                     0                     NA&
 10                       12%                     8                     2%
 11                       33%                     3                    16%
 12                       41%                     3                     3%
 13                       21%                     8                     4%
 14                       26%                     0                     NA&
 15                       19%                     3                     3%
 16                       13%                     9                     4%
 17                       50%                    11                     1%
 18                        7%                    20                     2%
 19                       32%                     3                     4%
 20                       61%                     1                     2%
 21                       24%                     6                     5%

 All                      33%                   146                     4%
* Total penetration is the sum of the filter penetration and face seal leakage.
  For example, a total penetration of 25% corresponds to an exposure equal to 1/4 of
  the exposure without a respirator. Total penetration is expected to be <=10% for 
  this class of respirators. 
+ Ninety-five percent of wearers are expected to have total respirator penetration less 
  than the stated value. For this class of respirators a value of <="10%" is expected. 
& This model failed the fit test (i.e., had a first-donning face fit leakage <="1%)" 
  with all 25 persons; therefore, the 95th percentile total penetration could not be 

Return to top.

Disclaimer   All MMWR HTML versions of articles are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the electronic PDF version and/or the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.

**Questions or messages regarding errors in formatting should be addressed to

Page converted: 12/14/98


Safer, Healthier People

Morbidity and Mortality Weekly Report
Centers for Disease Control and Prevention
1600 Clifton Rd, MailStop E-90, Atlanta, GA 30333, U.S.A


Department of Health
and Human Services

This page last reviewed 5/2/01