The article deals with the mechanisms of generation and functioning of spontaneous speech and their influence on the interpreter’s spontaneous speech during simultaneous interpreting. The subject of the research is spontaneous speech in simultaneous interpretation, the object of research is functional notional types of oral speech which serve as the foundation for the algorithm of expressing the main idea of the message. Depending on the functional notional type of spontaneous speech that can be considered as the dominant one in the source message, the interpreter faces different challenges. Following the original message too closely may lead to an abundance of self-corrections, repetitions, slips of the tongue in the interpreter’s spontaneous speech, which, together with confusing syntactic constructions, make it difficult for the audience to understand the target message. Interpreting simultaneously at a higher cognitive level within the model of comparative cognitive transformations, which takes into account algorithms of expressing the main idea, the interpreter can produce a more coherent and smooth spontaneous speech in the target language based on the cognitive essence as a result of thought transformation of the original message. Pragmatic orientation of spontaneous speech predetermines methods of compression as a tool of pragmatic adaptation of the original message within the cognitive pragmatic circle of oral communication, which can be extrapolated to different functional styles of speech and types of discourse.