Purpose It is not easy to assess how severe and annoying a patient's snoring is. Solid parameters are lacking; snorers cannot deliver a reliable self-assessment and it is uncertain whether bed partners' statements can be relied upon. The purpose of the present study was therefore to investigate whether and how well snoring assessment based on acoustic parameters and bed partners' reporting agree. Methods In a double-blind, placebo-controlled study on snoring treatment, several acoustic parameters [snoring index (SI), percentage snoring time (ST), sound pressure level, sound energy, loudness, psychoacoustic annoyance and psychoacoustic snore score (PSS)] were measured in 18 subjects during 24 polysomnographies. Bed partners also assessed snoring annoyance and loudness as well as treatment outcome. Results No correlation was found between the subjective annoyance caused by snoring and the acoustic parameters. Regarding perceived loudness, there was a moderate, significant correlation with loudness (N-5) and PSS over the hour with the highest SI. SI, ST, LAeq and maximum sound pressure level dB(A)(max) showed no significant correlation. After the intervention only mean sound energy LAeq over the entire night showed a significant correlation (r(s) = 0.782; p = 0.022) with bed partners' assessments. However, this result was not confirmed in the second control night. Conclusions The non-existent or only weak correlation between bed partners' ratings and objective parameters indicate that snoring severity should be evaluated with caution. Neither acoustic parameters, at least for one measurement over just one night, nor bed partners' ratings should be used as the sole basis for snoring assessment.