Background. Tears of the rotator cuff are common, some are repaired surgically. There are problems associated with repair and arthroscopic knot tying, and retear rates can be as high as 94%. It is known that surgical knots can slip or break and may cause ischemia and infection, affecting the healing process. Method. A barbed suture was developed. A materials testing machine measured the pull out strength of the sutures from the ovine infraspinatus tendon. Four groups of eight tendons were tested, using one, two, three or four knot-free suture loops for each repair. The maximum tensile strength was recorded. A control group with one knotted suture was used. Results. One suture had a mean strength of 38.9 N, two sutures 82.3 N, three sutures 138.8 N, and four sutures 176.6 N; each difference was significant. The single knotted suture of the same material had a mean strength of 59 N. Discussion. Barbed sutures inserted without knots had significant tensile strength. They may offer an alternative to current rotator cuff repair methods, being technically easier and quicker. Conclusion. The use of barbed sutures may be feasible in the repair of the rotator cuff, but further testing is required.