Article Text

Download PDFPDF

Sexual behaviour and risk reduction strategies among a multinational sample of women who have sex with women
  1. Vanessa Schick1,
  2. Joshua G Rosenberger2,
  3. Debby Herbenick1,
  4. Michael Reece1
  1. 1Center for Sexual Health Promotion, Indiana University, Bloomington, Indiana, USA
  2. 2Department of Global and Community Health, George Mason University, Fairfax, Virginia, USA
  1. Correspondence to Dr Vanessa Schick, Center for Sexual Health Promotion Indiana, University, 1025 E. 7th St., HPER 116, Bloomington, IN 47405, USA; vschick{at}


Objectives The development of safer sex recommendations for women who have sex with women (WSW) remains challenging given a limited understanding of sexual behaviour between women. The present study was conducted in order to investigate the sexual repertoires of WSW and the safer sex methods they use to reduce the likelihood of sexually transmitted infection acquisition.

Methods An online survey targeted towards women with desire, attraction or previous sexual behaviour with women was distributed globally. Women (N=3116) who engaged in at least one sexual act with a woman in the previous year and were currently living in the USA, UK, Canada or Australia were included in the present study. Questions were based upon previously validated items in nationally representative studies.

Results Participants indicated a wide diversity of sexual behaviours with the majority of women reporting a history of genital rubbing (99.8%), vaginal fingering (99.2%), genital scissoring (90.8%), cunnilingus (98.8%) and vibrator use (74.1%). Barrier use was reported by a minority (<25%) of the participants.

Conclusions The variety of sexual acts reported by the sample points to the need for the development of more contextually appropriate sexual health guidelines for WSW.

  • Sexual behaviour
  • women who have sex with women
  • WSW
  • risk reduction
  • bisexual
  • sexual experience
  • sexual health
  • lesbians
  • gay men
  • homosexuality
  • public health
  • sexuality
  • social

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Though nationally representative studies in the USA estimate that between 5% and 16.8% of women in some age groups have engaged in sexual behaviour with women,1–3 little remains known about the sexual lives of women who have sex with women (WSW). Understanding the sexual health needs of WSW is important because researchers have documented the transmission of trichomoniasis,4 genital herpes,3 human papillomavirus5 ,6 and HIV between women.7–10 Furthermore, WSW may be at a higher risk for bacterial vaginosis than other women.8 ,11–14 Yet, knowledge regarding the behavioural mode of transmission has been hindered by the absence of data regarding the sexual repertoires of WSW.15 ,16

Several nationally representative studies have assessed the prevalence of sexual behaviours between women including receptive/active oral intercourse, genital contact and vibrator/dildo use.2 ,17 While, together, these data are useful in understanding the prevalence of select activities between women, the diversity of sexual behaviours occurring between women remains unknown. Additionally, generalisations about behaviour from minority populations in population-level studies are limited by their small sample sizes. In contrast, several studies have concentrated on sexual behaviour among large samples of WSW. Although these studies were successful in documenting the existence of behavioural diversity between women, they assessed behaviours using subjective measurements (eg, ‘commonly’, ‘frequently’).18 ,19 This is problematic as participants may vary in their operationalisation of ‘frequent’ or ‘commonplace’ behaviour. Furthermore, research on sex toy use among WSW has not distinguished between the specific types of products used, the way(s) in which women were using the toy and the type/gender of partner they were using the toy with.18–21

The methodological drawbacks and lack of detailed sexual behaviour data between women have challenged the development of safer sex recommendations for WSW. To that end, risk estimates and corresponding sexual safety recommendations for WSW are often modified from those provided to heterosexual dyads, emphasising the use of barriers during cunnilingus and toy use. However, the likelihood of contracting certain sexually transmitted infections (STI) via cunnilingus may be less likely than transmission through other sexual behaviours that include genital-on-genital contact as such behaviours provide the means for transmission of infected cervicovaginal secretions.15 Also, using a toy without a barrier presents little to no risk for STI transmission if the toy is not shared, is adequately cleaned between uses and/or is used with a monogamous partner who does not have transmissible infections. Accordingly, the design and implementation of informed public health programmes targeted at reducing STI transmission between women is contingent upon the availability of reliable data on the sexual activities that women engage in, the context in which they engage in them and the measures they take to protect themselves while engaging in them.


The purpose of the present study was to describe women's sexual behaviour history with other women in order to assess common sexual behaviours engaged in by WSW. Predictors of sexual behaviour including sociodemographic and relational/contextual differences were evaluated in order to improve understanding about the diverse needs of various groups of WSW. In addition, the prevalence of safer sex strategies described by WSW were assessed.


Participant recruitment and data collection

To facilitate successful recruitment of a diverse group of participants, sampling approaches were duplicated from those used in recent studies of sexual behaviour among other traditionally hard-to-reach populations.22 ,23 Recruitment messages were distributed globally through online/paper media, website discussion forums, social networking websites and various websites/blogs over several months from 2010 to 2011. Additionally, the study was conducted via a participatory effort with a European-based website with a large membership of women who were interested in connecting with other women. Participants aged 18 years and over who spoke English were invited to participate if they had experienced or currently experienced sexual attraction, desire, affection, intentions towards or sexual behaviour with other women.

The recruitment message included a summary of the study. Participants who anonymously consented were directed to the survey which took approximately 20 minutes to complete. Upon completion, participants were provided with an electronic coupon for use at an online retailer. All study protocols were reviewed and approved by the Institutional Review Board at the academic institution of the first author.

Main outcome measures

Participants completed a variety of measures regarding their sociodemographic characteristics, sexual behaviour and safer sex practices. Sociodemographic measures included country of current residence, age, gender, sexual orientation, race, ethnicity and level of education completed (table 1). Participants were also asked about the duration of their current relationship status (dichotomised as over or under 1 year) and number of non-relationship sexual partners (none, one, more than one).

Table 1

Participant sociodemographic characteristics by mean number of sexual acts in the past year

Participants' sexual behaviour history was assessed using questions regarding the timeline of their sexual behaviours (see table 2 for a description of the behaviours). Response options and language (when appropriate) for the toy use questions were adapted from a US national study on vibrator use17 ,24–26, and the sexual behaviour questions were adopted from the National Survey of Sexual Health and Behaviour (NSSHB).1 ,27 Behaviours that were not assessed in these earlier studies were assessed using the same response options as the other sexual behaviours (‘Past Month’, ‘Past 3 Months’, ‘Past Year’, ‘More than a Year Ago’ and ‘Never Done This’). Participants who indicated that they had shared a sex toy with a sexual partner were asked about the nature of their relationship with that partner status (‘monogamous relationship’, ‘non-monogamous relationship’, ‘unsure’). Participants were then asked about the behaviours that occurred during the most recent time they shared the respective toys including whether they ‘placed the toy on their partner's genitals’ and/or ‘inserted it in their own/their partners' vagina’. Those who reported use on their own and their partner's genitals were asked about their sexual safety methods. Specifically, they were asked whether they ‘cleaned the vibrator/dildo before using it’, ‘cleaned the vibrator/dildo after using it’, ‘put a condom over the vibrator/dildo before using it’ and/or ‘changed the condom over the vibrator/dildo if both you and your partner inserted it’.

Table 2

Types of sexual behaviour engaged in by participants who reported sexual behaviour with another female partner within the past year


Female participants were included in the present study if they reported having at least one female sexual partner in the previous year. Participants were excluded from analyses if they reported their current country of residence as other than the USA, UK, Canada or Australia. These geographical locations were chosen because they represented the majority (93.9%) of the sample and because of similarities in the language and customs between countries. All other individual countries represented <1% of the sample.

While a quarter of the sample reported engaging in sexual behaviour with a man in the past year (25.4%, n=791), the purpose of the present study was to assess the sexual behaviours that women engage in with other women. Thus, only frequencies of various sexual behaviours and precautionary behaviours with women were assessed. Participants were stratified using similar age categories as those presented in the NSSHB.1 An ANOVA was conducted to investigate mean differences in the number of sexual behaviours engaged in by WSW as a function of various sociodemographic characteristics. A bivariate logistic regression model was conducted to assess the relationship between age, relationship status and current country of residence to specific sexual behaviours.


Sociodemographic characteristics

Most of the sample was from the US (44.7%, n=1390) or the UK (49.7%, n=1546) with a minority of the sample reporting their current residence as Canada or Australia (see table 1). Participants ranged in age from 18 to 69 with an average age of 29.27 (SD=9.10; median=27). Most participants were well educated, White and identified as a lesbian. Less than a quarter of participants indicated that they had currently had sexual partners outside of a monogamous relationship (24.7%, n=748).

The mean number of sexual acts that the participants reported engaging in with another female in the past year varied as a function of age, education, race and sexual/relationship status (represented in table 1). Participants in the 18–23 age cohort reported significantly fewer sexual activities in the past year than participants in the 24–49 age range. Accordingly, participants with a high education degree reported significantly more activities than those with a high school degree. Those with more than one current sexual partner reported the highest number of sexual activities.

Sexual behaviour

Most participants reported genital rubbing (99.8%), vaginal fingering (99.2%), vaginal fisting (56.7%), cunnilingus (98.8%) and genital scissoring (90.8%) at least once. Of those who indicated that they had participated in each behaviour, at least half reported that they engaged in the behaviour within the past 30 days (see table 2). The most commonly reported behaviour was genital rubbing and the least common was vaginal fisting. The percentage of participants who indicated sex toy use with a female partner varied by toy type. While most participants indicated that they had used a vibrator (74.1%, n=2279), non-vibrating strap-on (56.5%, n=1736) or non-vibrating dildo (55.6%, n=1709) at least once in their lifetime with a female partner, fewer participants indicated that they had used a non-vibrating double dildo (22.8%, n=702), butt plug (10.5%, n=323) or anal beads (6.6%, n=203) in their lifetime. Approximately a quarter of the participants reported that they had used a vibrator, non-vibrating dildo or non-vibrating strap-on during the past 30 days (see table 2). Of those participants, 77.2% (n=611) reported that they inserted the vibrator vaginally and 14.4% (n=114) reported that they inserted the vibrator anally. Additionally, 95.1% (n=694) of the participants reported inserting the dildo/butt plug/anal bead vaginally and 22.2% (n=568) reported inserting it anally.

When age, country of residence and relationship status were entered into a multivariate model predicting sexual behaviour over the past year, both age and relationship status predicted multiple sexual behaviours (table 3). Younger participants were more likely to report that they had engaged in genital rubbing, vaginal fingering and/or cunnilingus during the past year. In contrast, older age was predictive of vaginal fisting; genital scissoring and using a vibrator, non-vibrating dildo, butt plug or anal beads in the past year. Being in a relationship for over a year consistently increased the odds that participants would engage in sexual behaviours in the past year.

Table 3

Sexual behaviour in the past year stratified by age and relationship status

Sexual safety methods

Participants were asked about their use of several sexual safety methods. The majority of participants reported having shared a vibrator (58.2%, n=1441) and/or dildo (59.3%, n=1122) with a sexual partner in their lifetime. Of those, participants who reported using the toy on both their own and their partner's genitals were asked whether they cleaned the toy and/or used a barrier over the toy during the most recent time they shared a toy. Regardless of whether the participant was in a relationship at the time of the event, almost three-quarters of all participants reported cleaning the vibrator and dildo before the event and over 80% reported that they cleaned the toy after the event. However, a minority of women reported that they used a condom on the toy at any time during the event. Fewer women who were not in a relationship (11.9%, n=12) reported that they used a condom on their shared vibrator than the percentage of women who were in a relationship (17.3%, n=117). This finding was reversed for dildo use with more non-relationship women reporting that they used a condom on their dildo (33.8%, n=45). Less than a third of participants reported changing the condom on their shared vibrator and/or dildo (presented in table 4).

Table 4

Participation in safer sex strategies with a female partner by relationship status


Participants in this sample of WSW reported a large variety of sexual behaviours with the majority of participants reporting a lifetime history of genital rubbing, vaginal fingering, genital scissoring and vibrator use.19 The diversity of women's sexual repertoires with other women suggests that modifying heterosexual models of risk to construct WSW sexual safety recommendations may fail to equip WSW with the necessary information to reduce their likelihood of STI transmission. Consistent with findings from previous research, over nine of 10 participants in the present study reported engaging in genital scissoring.18 ,19 Of those participants, almost 95% indicated that they never used a barrier during this sexual act. Additionally, while the majority of participants reported that they never shared a vibrator or dildo with a female sexual partner, only a small percentage of the participants reported using/changing a barrier over the toy during the sexual encounter. Although it is not clear from this study why most participants did not use barriers during genital scissoring and dildo/vibrator toy use, other studies have found that WSW use few, if any, safer sex practices because they are not at risk for pregnancy with a female partner and they perceive sexual behaviours with WSW to pose little risk for STI transmission.15 ,16 ,20 ,28–30 However, the documented transmission of trichomoniasis,4 genital herpes3 and human papillomavirus5 ,6 between women7–10 suggest that these behaviours may pose a risk of STI transmission. WSW should be made aware of these risks, so that they can make informed decisions about their sexual lives.

Perceptions that WSW are not at risk for STI may be fuelled by the lack of public health messages coupled with the limited availability of products designed specifically to reduce STI transmission between women.31 In the same way that sexual safety recommendations for WSW are modified from those designed for WSM, women who choose to use a latex barrier to reduce the risk of STI transmission during genital scissoring or toy use with another woman are limited to products that require modification from their original Food and Drug Administration-approved purpose (penile–vaginal intercourse).32 As a result, women who choose to use a latex barrier during genital scissoring have to modify a condom or seek out and lubricate a dental dam. Similarly, because male condoms are designed for use during penile–vaginal intercourse, they are shaped to fit on a phallic-shaped object. While the shapes and sizes of the sex toys used by the participants remains unknown, it is unlikely that all of the toys used were phallic shaped given the vast assortment of shapes and sizes of various sex toys.26 Instead of recommending male condoms, the invention and promotion of products targeted specifically towards WSW may encourage safer sex between women by (1) acknowledging and validating that sexual behaviour between women poses a risk and (2) providing the subsequent means necessary to reduce the barriers WSW experience when they choose to protect themselves against such risks. Future research should investigate whether the distribution of safer sex products targeted towards WSW does, in fact, modify risk perceptions among WSW.

Given the number of modifications that would be necessary in order to use a male condom during certain sexual behaviours, WSW may also benefit from the promotion of alternative safer sex methods. In addition to cleaning the toy before/after each use, certain types of toys may decrease the likelihood of STI transmission. Of the women who reported using a dildo with a female partner, the lowest percentage of women reported that they used a double-dildo. A barrier may not be needed with a double-dildo if, during partnered use, both women consistently use their own clean end of the dildo. However, use should be considered within the larger context of the sexual encounter as the double-dildo may increase the likelihood of genital–genital contact if both women are using the dildo simultaneously. Further comprehensive research focused specifically on WSW sexual behaviour at an event level is necessary in order to inform assessments of the benefits/risks of various sexual behaviours.

In addition to STI risks that may occur from exchanging infected cervicovaginal secretions and skin-to-skin contact during toy use and genital scissoring, approximately half of the participants reported a history of vaginal fisting. Depending on the technique used and force exerted, fisting may result in vaginal tearing.20 Women who participate in vaginal fisting (and other types of digital vaginal penetration) may benefit from guidance regarding nail hygiene, hand washing, glove use and/or the use of lubricants to minimise their risk of vaginal tearing and STI/HIV transmission and maximise their overall sexual experience.31

Although this study is the first to document the existence of a wide diversity of sexual behaviours in a global sample of recently sexually active WSW, certain behaviours (eg, analingus, sadomasochistic practices) were not accounted for in this study. Similarly, the inclusion of only recently sexually active WSW failed to take into consideration the experiences of those women who opted not to engage in sexual behaviour with a female partner. Additionally, because the survey was distributed online, there is a possibility that all participants were not unique. However, efforts were made to reduce the likelihood that participants would take the survey more than once (ie, the survey could only be accessed once on a computer, the participation incentive was not cumulative). Finally, while this sample was large, it may not be representative of all WSW who have had recent sexual activity as the majority of the women were recruited online, and all participants completed the questionnaire on the internet. Consequently, eliminating all WSW without access to the internet may have limited the diversity of the sample. Furthermore, participants were recruited online from websites that may be more appealing to women who are more sexually conscious. However, this limitation is also a strength in that previous studies have demonstrated online data collection has the potential to limit social desirability bias and result in more accurate and honest responses from study participants.33 ,34 While challenges exist with establishing truly nationally representative samples of WSW, the similarity of the reported behaviour rates (eg, vibrator/dildo use) to other representative samples suggests that the behaviours of the women in this sample may be similar to those of the women in more representative samples of WSW.17 However, future research should continue to explore the diversity of women's behaviours with other women in assorted samples.


The majority of WSW in this sample reported a diverse sexual repertoire with almost all women reporting a history of genital rubbing, vaginal fingering, cunnilingus, genital scissoring and vibrator use. Few participants reported taking safety precautions during various sexual behaviours. Thus, WSW may benefit from the distribution of sexual risk reduction strategies targeted specifically at their sexual health needs.

Key messages

  • Over three-quarters of the sample reported a history of genital rubbing, vaginal fingering, genital scissoring and cunnilingus within the past year.

  • Younger participants reported fewer behaviours that included vaginal/anal insertion than older participants. Participants in long-term relationships reported a wider diversity of behaviours.

  • Barrier use while sharing a vibrator (17%), dildo (21.0%) or engaging in genital scissoring (5%) was not commonly reported by participants.

  • Targeted interventions should focus on the diverse sexual behaviours that may put WSW at risk.


This study was conducted in partnership with GaydarGirls. The authors would also like to thank Searah Deysach for her inkind support of this research and Chelsea Heaven for her assistance with database construction. The authors would also like to thank Searah Deysach for her in-kind support of this research, Chelsea Heaven for her assistance with database construction and J. Dennis Fortenberry for ongoing advice/guidance.



  • Funding Vanessa Schick was partially funded by a Developmental Award granted by the American Sexually Transmitted Diseases Association.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by Indiana University-Bloomington.

  • Provenance and peer review Not commissioned; externally peer reviewed.