Trophic ecology has benefitted from the use of stable isotopes for the last three decades. However, during the last 10 years, there has been a growing awareness of the isotopic biases associated with some pre-analytical procedures that can seriously hamper the interpretation of food webs. We have assessed the extent of such biases by: (1) reviewing the literature on the topic, and (2) compiling C and N isotopic values of marine invertebrates reported in the literature with the associated sample preparation protocols. The factors considered were: acid-washing, distilled water rinsing (DWR), sample type (whole individuals or pieces of soft tissues), lipid content, and gut contents. Two-level ANOVA revealed overall large and highly significant effects of acidification for both delta(13)C values (up to 0.9 per thousand decrease) and delta(15) N values (up to 2.1 per thousand decrease in whole individual samples, and up to 1.1 per thousand increase in tissue samples). DWR showed a weak overall effect with delta(13)C increments of 0.6 per thousand (for the entire data set) or decrements of 0.7 per thousand in delta(15) N values (for tissue samples). Gut contents showed no overall significant effect, whereas lipid extraction resulted in the greatest biases in both isotopic signatures (delta(13)C, up to -2.0 per thousand in whole individuals; delta(15)N, up to +4.3 per thousand in tissue samples). The study analyzed separately the effects of the various factors in different taxonomic groups and revealed a very high diversity in the extent and direction of the effects. Maxillopoda, Gastropoda, and Polychaeta were the classes that showed the largest isotopic shifts associated with sample preparation. Guidelines for the standardization of sample preparation protocols for isotopic analysis are proposed both for large and small marine invertebrates. Broadly, these guidelines recommend: (1) avoiding both acid washing and DWR, and (2) performing lipid extraction and gut evacuation in most cases.