We evaluated the long-term benefits of steroid injection in 20 consecutive patients (25 shoulders) with primary acromioclavicular arthritis using the Constant score. All patients were followed up for a minimum of 5 years. The average age of the patients was 55 years. The mean pre-injection Constant score was 61 points. At six months this improved to 81 points, (mean difference – 19.36) which was highly statistically significant (p<
0.01). There was further improvement at 12 months, on the 6-month score, mean 86 points that also showed a significant improvement on the 6-month score (p = 0.001). At 5 years the mean score was 81 points and this was a significant detoriation compared with the 12-month score (p=0.01) but still a significant improvement when compared to the preinjection scores (p<
0.0005). In addition, the younger the patients, the greater the improvement in the objective score which measures range of movement and power (r= − 0.47; p= 0.01). Female patients also had a greater improvement (r=0.405; p=0.05). We conclude that local steroid injection is an effective method of treatment for primary isolated acromioclavicular arthritis and improvement continues for at least 12 months and the benefit is felt up to 5 years though the pain relief tends to tail off long-term.