Background: Despite the widely reported success of knee arthroplasty, studies show that 1.6–3 % of patients undergo revision within the first postoperative year predominantly due to infection. Preoperative skin preparation may potentially decrease the bacterial load and consequently, the risk of periprosthetic joint infections. The effects of hair removal on prosthetic joint infection are inconsistent. Our primary aim was to investigate if hair removal with a clipper influenced skin colonisation and bacterial composition. Methods: Forty Caucasian male participants who were planned to undergo knee arthroplasty, (mean age 63.8 years), were included. Patients were randomised to hair removal in a within-person study design. As a control, the opposite leg of the patient was used. Swabs were collected prior to hair removal (baseline), immediately after hair removal (Day 0), and with follow-up after one and seven days. Results: The intervention showed significant decrease in mean log colony-forming units per. cm² from baseline 2.97 to 2.67 (P<0.01) immediately after hair removal and sustained at Day 1 (P=0.01). At Day 7, the mean was non-significant compared to baseline. The control group did not show any decrease of skin microbiota at follow-up on Day 0, 1 or 7. No significant differences within the bacterial composition were found between the intervention and control leg at baseline among the six most prevalent detected bacterial species: Staphylococcus epidermidis, Micrococcus luteus, S. hominis, S. capitis, S. haemolyticus and S. aureus. The study did not find any changes in the bacterial composition over time. Conclusion: Hair removal with a clipper within 24 hours prior to surgery causes a significant non-selective reduction in skin colonisation.