Over recent years, revision stems have been introduced for the treatment of periprosthetic fractures. Modular stems are supposed to provide some advantages to the surgical treatment of this complication. To evaluate the midterm radiological and functional results of periprosthetic hip fractures treated with modular revision stems. A retrospective review of 21 patients with a mean age of 77 years (47–88) is here presented. They had suffered a periprosthetic femoral fracture around a total hip arthroplasty between 2004 and 2010. Sixteen of them were a Vancouver classification type B2 and 5 were type B3. In all cases, a transfemoral approach was performed. Previous stems were exchanged for Revitan® (Zimmer) modular stems. There were 14 curved stems and 7 straight stems. The mean patient follow-up period was 22 months. Functional results were assessed using the Merle d’Aubigne score. The degree of subsidence was determined by comparing the postoperative images with those taken during the very last follow-up. No stem loosening was observed during follow-up. There was one non-union of the femoral osteotomy which was not of either clinical or functional relevance. Subsidence was less than 4 mm in 19 cases; one case showed 7 mm of subsidence and the other had 21 mm at final follow-up. Three patients (14 %) had to be operated on again. One of them was due to a refracture, another for acetabular loosening and one more due to reiterative dislocation (an additional case of dislocation was conservatively treated). The mean Merle d’Aubigne score at final follow-up was 5-5-4. Mortality stood at 19 % throughout follow-up. Modular revision stems provide some advantages compared with osteosynthesis in that they make early weight bearing possible. In addition, version, offset, and limb length is easier to control and calculate. Therefore, modular revision stems should be considered the treatment of choice for periprosthetic femoral fractures.