A phenomenon initially reported by Chen and Vinther in 2013 [1], and now commonly referred to as low endotoxin recovery (LER), has prompted the Food and Drug Administration (FDA) to request specific data demonstrating the capability of the LAL BET method (i.e., USP <85>) to recover endotoxin from spiked samples over time. The results of these spike/hold recovery studies are expected to be included in the Biologics License Applications (BLA) for review by the Center for Drug Evaluation and Research (CDER) Hughes (2014) and Hughes et al. (2015) [2,3]. Such studies involve spiking a known amount of a surrogate endotoxin, such as purified lipopolysaccharide (LPS), into undiluted biological products and then testing at different time points to determine the recovery over time. We report here the experience and learning gained from conducting spike/hold recovery studies for a monoclonal antibody (Mab) product. Results from initial hold studies spiked with purified LPS showed rapid loss of endotoxin activity in the drug substance (DS) and significant batch-to-batch variation in the drug product (DP). After careful review and examination of the experimental details, it was determined that the study design and execution differed from the routine batch release USP <85> BET method with regard to mixing time and sampling scheme. The hold study design was subsequently revised so that the mixing time and sampling were the same as the verified USP <85> BET method used for routine batch release testing. The spike/hold recovery studies were repeated and the results demonstrated that LPS could be consistently recovered over time. These findings highlight the importance of carefully controlling sample preparation procedures in a spike/hold recovery study in order to demonstrate the suitability of using the LAL BET method for endotoxin detection.