STUDY QUESTION: How do the demographic characteristics, motivations, experiences and expectations of unregulated sperm donors (men donating sperm online through a connection website) compare to sperm donors in the regulated sector (men donating through a registered UK sperm bank)? SUMMARY ANSWER: Online donors were more likely to be older, married and have children of their own than sperm bank donors, were more varied in their preferences and expectations of sperm donation, and had more concerns about being a sperm donor. WHAT IS KNOWN ALREADY: While studies have examined motivations and experiences of both regulated sperm bank, and unregulated online sperm donors, no study has directly compared these two groups of donors. STUDY DESIGN, SIZE, DURATION: An email was sent to the 576 men who were registered sperm donors at the London Sperm Bank, the UK's largest sperm bank regulated by the Human Fertilisation and Embryology Authority (HFEA), who had commenced donation between January 2010 and December 2016, and had consented to be contacted for research. The online survey, which contained multiple choice and open-ended questions, was completed by 168 men over a 7-week period. The responses were compared to those of sperm donors registered on Pride Angel, a large UK-based connection website for donors and recipients of sperm: our research team had already collected these data. In total, 5299 sperm donors were on Pride Angel at time of data capture and 400 men had completed a similar survey. The responses of 70 actual online sperm donors (i.e. those whose sperm had been used to conceive at least one child) were used for comparison with the sperm bank donors. PARTICIPANTS/MATERIALS, SETTING, METHODS: The survey obtained data on the sperm donors' demographic characteristics, motivations, experiences and expectations of sperm donation. Data from sperm bank donors were compared to online donors to examine differences between the two groups. The study compared online and clinic donors who had all been accepted as sperm donors: online donors who had been 'vetted' by recipients and sperm bank donors who had passed the rigorous screening criteria set by the clinic. MAIN RESULTS AND THE ROLE OF CHANCE: A response rate of 29% was obtained from the sperm bank donors. Online donors were significantly older than sperm bank donors (mean ± SD: 38.7 ± 8.4 versus 32.9 ± 6.8 years, respectively) and were more likely to have their own children (p < 0.001 for both characteristics). Both groups rated the motivation 'I want to help others' as very important. Online donors rated 'I don't want to have children myself', 'to have children/procreate' and 'to enable others to enjoy parenting as I have myself' as more important than sperm bank donors, whereas sperm bank donors rated financial payment as more important than online donors, as well as confirmation of own fertility. Most (93.9%) online donors had donated their sperm elsewhere, through other connection sites, fertility clinics, sperm banks or friends and family, compared to only 2.4% of sperm bank donors (p < 0.001). There was a significant difference in how donors viewed their relationship to the child, with online donors much less likely than sperm bank donors to see their relationship as a 'genetic relationship only'. Online donors had more concerns about being a donor (p < 0.001), for example, being concerned about 'legal uncertainty and child financial support' and 'future contact and uncertainty about relationship with donor-conceived child'. LIMITATIONS, REASONS FOR CAUTION: Findings may not be representative of all sperm donors as only one online connection site and one HFEA registered sperm bank were used for recruitment. WIDER IMPLICATIONS OF THE FINDINGS: Despite concern regarding shortages of sperm donors in licensed clinics and unease regarding the growing popularity of unregulated connection websites, this is the first study to directly compare online and sperm bank donors. It highlights the importance of considering ways to incorporate unregulated online sperm donors into the regulated sector. With many online donors well aware of the legal risks they undertake when donating in the unregulated online market, this would both increase the number of sperm donors available at clinics but also provide legal protection and support for donors. STUDY FUNDING/COMPETING INTEREST(S): This study was supported by the Wellcome Trust Grants 104 385/Z/14/Z and 097857/Z/11/Z. The authors have no conflicts of interest.