We evaluated commonly used methods for monitoring stream restorations to inform and improve restoration monitoring and evaluation, using a headwater stream in the Oregon Coast Range as a case-study example. In-stream restoration projects are seldom monitored both pre- and post-restoration. In addition, frequently used low-cost methods may not provide sufficient data to effectively assess trends in stream temperature. Here, we examined what can be learned from temperature loggers installed in the same locations over multiple years in a restored stream. In-stream structures were installed between 2007 and 2011 along a 10-km length of South Sister Creek, Oregon for the purpose of enhancing in-stream habitat. Summer stream temperature data were collected using Hobo Pro-V temperature-logging thermistors at four locations in 2006, prior to restoration, as well as in 2012 and 2013. In 2013, additional temperature loggers were placed within 80 m of the four original loggers to investigate within-reach variability. Although median stream temperatures were highest in 2013 at all four multi-year sites, 7-day maximum temperatures were 4 to 5 °C cooler in post-treatment years than in 2006. Inter-annual variability in stream temperature was more closely linked to inter-annual variability in air temperature and solar radiation than presence of in-stream structures. Thermal heterogeneity was greater in reaches with deeper pools than in bedrock-dominated reaches. Although in-stream structures can create cool microhabitats, they have little influence on mean stream temperature. Longer pre-restoration monitoring and sensor deployment in more varied in-stream habitats would improve our ability to evaluate restoration impacts.