Objectives: This study evaluated the percentages of continuous margins (%CM) and fracture strength (FS) of crowns made out from blocs of leucite-reinforced ceramic (IPS Empress CAD) and luted with a representative self-adhesive cement (RelyX Unicem) under four contaminating agents: saliva, water, blood, a haemostatic solution containing aluminium chloride (pH= 0.8) and a control group with no contamination. Study Design: %CM at both tooth-cement (TC) and cement-crown (CC) interfaces were determined before and after a fatigue test consisting of 600'000 chewing loads and 1'500 temperature cycles changing from 5º C to 50º C. Load to fracture was recorded on fatigued specimens. Kruskal-Wallis test was used to compare %CM and FS between the five groups with a level of confidence of 95%. Results: At the TC interface, no significant differences in marginal adaptation before loading could be detected between groups. After loading, a significant marginal degradation was observed in the group contaminated with aluminium chloride (52 ± 22 %CM) in respect to the other groups. No significant differences in %CM could be detected between the groups contaminated with saliva, water, blood and the control. At the CC interface, no significant differences in marginal adaptation were observed between the groups. The FS on loaded specimens was around 1637N, with no significant differences between groups as well. Conclusions: An adverse interaction of the highly acidic haemostatic agent with either dentin or the self-adhesive cement could explain the specimens' marginal degradation. The self-adhesive cement tested in this study was no sensitive to moisture contamination either with saliva, water or blood.