Seventy-three thumbs in 67 women with osteoarthritis of the trapeziometacarpal joint were assessed prospectively. The patients were randomly allocated for treatment by either trapeziectomy alone (n = 26) or trapeziectomy with tendon interposition (n = 23) or ligament reconstruction (n = 24). A K-wire was passed through the thumb metacarpal base and across the trapezial space into the distal pole of the scaphoid in all procedures; this held the base of the metacarpal away from the scaphoid for 4 weeks. Standard thumb radiographs were used to calculate the trapezial space ratio before surgery and at 1-year follow-up examination. The trapezial space ratio decreased significantly from a preoperative mean of 0.40 (range, 0.20-0.56) to a mean of 0.18 (range, 0.08-0.30) after trapeziectomy, 0.16 (range, 0.07-0.30) after trapeziectomy with tendon interposition, and 0.20 (range, 0.00-0.33) after trapeziectomy with ligament reconstruction. There was no significant difference between these 1-year follow-up trapezial space ratios, suggesting that the placement of a K-wire across the trapezial void is as effective as tendon interposition or ligament reconstruction in creating a trapezial space in the short term at least. However, the need to create a trapezial space and maintain thumb length is questioned because the trapezial space height did not correlate with thumb strength at 1-year follow-up examination.