The loss of seagrass meadows is an increasing problem worldwide. The important role that these meadows play in coastal ecosystems has resulted in substantial attention to the development of seagrass restoration techniques. Here, I present long-term (up to 5 years) results of seagrass restoration off the coast of Adelaide, South Australia, where >5,000 ha of seagrass has been lost and where trials of traditional restoration techniques using seeds and transplants have failed due to high levels of sand and water movement. Hessian (burlap) sandbags were deployed bimonthly (with some interruptions) from November 2007 to November 2012 (a total of 24 deployments), with a mix of single- and double-layered bags, to provide a stable substrate for naturally occurring Amphibolis seedlings to recruit to. At the end of the study (January 2013), bags deployed in August 2009 had similar stem densities to those found in adjacent natural meadows (15.2 ± 1.4 (SE) vs 18.6 ± 2.5). Bags deployed in May 2008 and August 2011 had 12.8 ± 2.3 and 13.2 ± 2.2 stems, respectively. Furthermore, stem lengths on older bags were greater than those on natural meadows (42.1 ± 4.2 after 62 months vs 30.2 ± 1.5 cm). While there was some interannual variation in recruitment success, the strongest predictor of success was deployment month. Bags deployed outside the austral winter recruitment season did not retain the ability to catch a large number of recruits, indicating that any restoration using this technique will have to be undertaken between approximately May and August to maximize chances of success.