We hypothesized that ultrasound (US) guidance improves outcomes of corticosteroid injection of trochanteric bursitis. 40 patients with greater trochanteric pain syndrome defined by pain to palpation over the trochanteric bursa were randomized to injection with 5 ml of 1% lidocaine and 80 mg of methylprednisolone using (1) conventional anatomic landmark palpation guidance or (2) US guidance. Procedural pain (Visual Analogue Pain Scale), pain at outcome (2 weeks and 6 months), therapeutic duration, time-to-next intervention, and costs were determined. There were no complications in either group. Ultrasonography demonstrated that at least a 2-in (50.8 mm) needle was required to consistently reach the trochanteric bursa. Pain scores were similar at 2 weeks: US: 1.3 ± 1.9 cm; landmark: 2.2 ± 2.5 cm, 95% CI of difference: - 0.7 < 0.9 < 2.5, p = 0.14. At 6 months, US was superior: US: 3.9 ± 2.0 cm; landmark: 5.5 ± 2.6 cm, 95% CI of difference: 0.8 < 1.6 < 2.4, p = 0.036. However, therapeutic duration (US 4.7 ± 1.4 months; landmark 4.1 ± 2.9 months, 95% CI of difference - 2.2 < - 0.6 < 1.0, p = 0.48), and time-to-next intervention (US 8.7 ± 2.9 months; landmark 8.3 ± 3.8 months, 95% CI of difference - 2.8 < - 0.4 < 2.0, p = 0.62) were similar. Costs/patient/year was 43% greater with US (US $297 ± 99, landmark $207 ± 95; p = 0.017). US-guided and anatomic landmark injection of the trochanteric bursa have similar 2-week and 6-month outcomes; however, US guidance is considerably more expensive and less cost-effective. Anatomic landmark-guided injection remains the method of choice, but should be routinely performed using a sufficiently long needle [at least a 2 in (50.8 mm)]. US guidance should be reserved for extreme obesity or injection failure.