Fusarium head blight (FHB) is one of the most important diseases of durum, spring, and winter wheat in Canada. Growers rely on an integrated strategy to manage the disease, including fungicide application at the current recommendation of early to 50% anthesis (BBCH61–65). This study evaluated the effect of fungicide application timing and seeding rates of durum wheat on FHB. Field trials were carried out from 2016 to 2018 at three locations in Saskatchewan. Eight treatments of the metconazole fungicide Caramba were applied to durum seeded at 75 and 400 seeds m−2. The fungicide treatments consisted of a nontreated check, a treated check, and applications at BBCH59, BBCH61, BBCH65, BBCH69, and BBCH73 and a dual application treatment (BBCH61 followed by BBCH73). FHB index, proportion of Fusarium-damaged kernels (%FDK), deoxynivalenol (DON), grain protein content (GPC, %), and yield were evaluated. Seeding rates influenced all parameters. The high seeding rate had a higher yield and FHB index but lower DON and GPC than did the lower seeding rate. All fungicide treatments resulted in lower FHB index, DON, and %FDK than the nontreated check. Under FHB conducive conditions, all anthesis applications from BBCH61 to BBCH69 had a similar effect on FHB index, %FDK, DON, and yield, whereas in years with low disease severity, the application at BBCH65 had lower FHB index, %FDK, and DON relative to other single applications. The dual application (BBCH61 + 73) treatment resulted in similar FHB index levels, %FDK, and DON content as the BBCH65 application at all site years. Our results indicate that the window of fungicide application can be extended to the end of flowering when FHB risk is high.