Piscirickettsiosis is an endemic disease of farmed salmonids in Chile, and is the main cause of infectious disease-related losses in the Chilean aquaculture industry. Inconsistent results with the use of vaccines and antimicrobials have led producers and government to search for alternative control measures. Fallowing sites between fish production cycles has been mandated by the government authority, but the effectiveness of this practice for preventing this disease has not been assessed under field conditions. We used a discrete-time survival analysis model to evaluate the effect of the duration of the fallow period on the hazard of piscirickettsiosis during the first 24 weeks of the production cycle on farms between September 2009 and August 2015. We compared the hazard of piscirickettsia for production cycles on farms that did and did not report the disease immediately before fallowing in the preceding cycle. We controlled for potential confounders, including external infectious pressure from neighboring farms. Our results showed that for both Atlantic salmon and rainbow trout there was no significant difference between the hazard of SRS for farms that reported the disease in the previous cycle and the comparison group, when these farms fallowed for more than three months. Shorter fallow periods were only assessed in rainbow trout, and findings indicate that the hazard of piscirickettsiosis is higher for farms with a recent history of the disease than for farms without a recent history of the disease prior to fallowing. These results suggest that fallowing for three months may be adequate to lower the exposure of P. salmonis from one production cycle to the next; however, our study also suggests there is a high hazard of observing SRS on farms, regardless of fallowing. (C) 2017 Elsevier B.V. All rights reserved.